Full Organic Aqueous Battery Based on TEMPO Small Molecule with Millimeter-Thick Electrodes

Article abstract of DOI:10.1021/acs.chemmater.8b03282

Thick electrodes with sodium and even anion intercalation organic compounds integrated in a neutral-pH aqueous battery offer unique advantages in terms of round trip efficiency, environmental impact, and scalability for off- or on-grid renewable energy st

Full text of DOI:10.1021/acs.chemmater.8b03282

Subscriber access provided by TULANE UNIVERSITY
Article
Full organic aqueous battery based on TEMPO
small molecule with millimeter-thick electrodes
Sofia Perticarari, Elodie Grange, Tom Doizy, Yann Pellegrin, Eric Quarez, Kenichi Oyaizu, Antonio
Jesus Fernandez-Ropero, Dominique Guyomard, Philippe Poizot, Fabrice Odobel, and Joël Gaubicher
Chem. Mater., Just Accepted Manuscript • DOI: 10.1021/acs.chemmater.8b03282 • Publication Date (Web): 19 Jan 2019
Downloaded from http://pubs.acs.org on January 20, 2019
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 17
Chemistry of Materials
1
2
3
4
5
6
7
8
9
Full organic aqueous battery based on TEMPO small molecule with
millimeter-thick electrodes
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Sofia Perticarari†‡, Elodie Grange†, Tom Doizy†, Yann Pellegrin‡, Eric Quarez†, Kenichi Oyaizuξ,
Antonio Jesus Fernandez-Ropero†, Dominique Guyomard†, Philippe Poizot*†, Fabrice Odobel*‡, Joël
Gaubicher*†
†Institut des Matériaux Jean Rouxel (IMN), Université de Nantes, UMR CNRS 6502, 2 rue de la Houssinière, B.P. 32229,
44322 Nantes Cedex 3, France.
‡CEISAM (Chimie et Interdisciplinarité, Synthèse, Analyse, Modélisation), Université de Nantes, UMR CNRS 6230, 2 rue
de la Houssinière, B.P. 92208, 44322 Nantes Cedex 3, France.
ξDepartment of Applied Chemistry, Waseda University, Tokyo 169-8555, Japan.
ABSTRACT: Thick electrodes with sodium and even anions intercalation organic compounds integrated in a neutral-pH aqueous
battery offer unique advantages in terms of round trip efficiency, environmental impact, and scalability for off or in-grid renewable
energy storage. Herein we report on the first anion-rocking chair / dual-ion organic battery. The latter reaches 35 Wh/kgmaterials at C/8
rate. It shows remarkable cyclability and coulombic efficiency in a cheap and neutral NaClO4 electrolyte pouch cell with highly
loaded millimeter-thick electrodes (5 mAh/cm²). This achievement is based on a thorough study of a commercial (2,2,6,6-
tetramethylpiperidin-1-yl)oxyl (TEMPO) benzene derivative namely 4-hydroxy TEMPO benzoate and its naphthalene analog (4-
carboxy TEMPO naphthalate) as positive electrode materials, and a bipyridinium-naphthalene oligomer as the negative electrode.
Combined UV-Vis spectroelectrochemistry and operando XRD account for the much improved cyclability of the hydrophobic 4-
carboxy TEMPO naphthalate at the expense of a lower specific capacity. This trend is reversed in the case of the 4-hydroxy TEMPO
benzoate derivative. Results show kinetic limitations of the 4-hydroxy TEMPO benzoate are associated with the surrounding
composite electrode, while inner-grain ionic and/or electronic transports play a decisive role for the 4-carboxy TEMPO naphthalate.
Introduction
Aqueous batteries constitute a novel and promising technology
towards renewable energy storage as they are inherently safe,
minimize cost and environmental impact by comparison to
other battery technologies1–4, thus offering a low cost and
greener alternative to Li-ion systems. This approach is still
affiliated however, with energy densities below 50Wh/kgcell
.
Furthermore, the relatively narrow electrochemical window
available in aqueous media together with costs and abundance
issues makes even more challenging the development of
appropriate materials. The latter should possess optimal
potential as well as high chemical and electrochemical stability.
As a result, the choice of greener and more abundant organic
materials appear as a logical step. So far only two families of
organic compounds have been demonstrating encouraging
performance as positive electrode active materials in aqueous
media: i) the p-type conjugated tertiary amine PTPAm showing
interesting experimental capacity and good cyclability in 21 m
LiTFSI water-in-salt electrolyte5 and ii) the nitroxyl radicals,
more precisely the 2,2,6,6-tetramethylpiperidinyl-N-oxyl
(TEMPO) derivatives. The latter is the most studied p-type
redox systems in aqueous electrolyte6–11. TEMPO undergoes a
one electron oxidation process to give the oxoammonium
cation, the charge being counterbalanced by the uptake of one
anion from the electrolyte (Figure 1).
Figure 1. Schematic representation of the electrochemical
reaction of TEMPO derivatives.
Up to now, TEMPO derivatives in aqueous batteries have been
mainly developed in the form of radical polymers. In this
regard, Oyaizu and Nishide’s group have demonstrated that the
achievement of high molecular weight and proper volume swell
ratio is fundamental for the good performance of the polymer12.
They extensively investigated TEMPO’s derivatives6–10,12,13
,
highlighting the difficulties in obtaining proper water
compatible TEMPO polymer. As a result, the poly(2,2,6,6-
tetramethylpiperidinyloxy-4-yl acrylamide) (referred to as
PTAm) revealed the most interesting performance, enabling the
development of two anionic rocking chair batteries.
Noteworthy, until recently, electrodes were based on thin films
of these polymers (from 50 nm to 3.2 µm)12. However, because
low cost (≤ 100 USD per kWhinstalled) is a sine qua non condition
to insure the economic viability of electrochemical systems for
grid storage, aqueous batteries need to be built with ultra-thick
ACS Paragon Plus Environment

Chemistry of Materials
Page 2 of 17
(>mm) electrodes. In this context, the same group has recently solution was diluted with dichloromethane and was washed
1
2
3
4
5
6
7
8
proposed thick composite electrodes of 3 mAh/cm² by
hybridizing PTAm with a 3D self-assembled mesh of single-
walled carbon nanotubes (SWCNT)7. Even though the cost of
SWCNT can potentially be an issue, impressive performance
has been obtained with only 1 to 5 wt.% SWCNT. Although all
these approaches are effective in that they successfully convey
organic materials with interesting performance as positive
electrode material for neutral aqueous battery, a full cell with
thick organic based electrodes of several mAh/cm2 still remain to
be demonstrated to the best of our knowledge. A different
approach is proposed in the present work based on “small
TEMPO molecules”, rather than polymeric structures. We note
that an example has been recently reported in the field of
supercapacitor by Itoi et al. who adsorbed the HTB (1-
Piperidinyloxy,2,2,6,6-tetramethyl-4-(2-oxo-2-phenylethyl)) in
the mesoporosity of active carbon (AC)14. Thanks to the huge
contact area between the finely dispersed HTB molecules and
the conductive AC surface, fast redox reactions of HTB could
be achieved in aqueous 1 M H2SO4. Moreover, the strong
adsorption capability of AC counteracts the dissolution of HTB
in the electrolyte. On the contrary, it is worth pointing out that
this method is inherently restricted to the supercapacitor field
because of the high loading of conductive agent. The best
performance (81% of capacity retention after 2000 cycles at
1 A/g) were typically obtained with 70 wt.% AC. Importantly,
this kind of strategy, that is well known especially in the Li-S
battery field since the work of Nazar15, can only mitigate
dissolution but not suppress it. Here we report on two different
strategies that have been followed with the aim to produce a low
cost and safe aqueous battery including p-type “small TEMPO
molecules”. Our approach will be supported by in-depth study
of electrochemical behaviors as well as demonstration of
potentialities and limitations of some of these materials in full
cells of up to 5 mAh/cm² using a new bipyridinium-naphthalene
diimide oligomer as the negative electrode.
with water (2 ), a saturated aqueous solution of sodium
bicarbonate (1 ), and brine (1 ). The organic layer was dried
over sodium sulfate, concentrated in vacuum and purified by
silica gel column chromatography to afford the compounds as
an orange powder.
1
4CT-Benzene: (yield 75%) H NMR (300 MHz, DMSO-d6) δ
7.36 – 7.47 (m, 3H), 7.21 – 7.29 (m, 2H), 2.95 (tt, J = 12.9 Hz,
1H) 1.96-1.97 (m, 2H), 1.58 (t, J = 12.8 Hz, 2H), 1.12 (s, 6H),
1.09 (s, 6H). HRMS (ESI)+ m/z calculated for C16H22NO3Na
299.1497 [M+Na]+ found 299.1503. FTIR (cm-1): 2830-3000
(rocking CH2), 1744 (C=O ester), 1366 (N-O. stretch).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4CT-Napthalene: (yield 80%) 1H NMR (300 MHz, DMSO-d6)
δ 7.94 (ddd, J = 12.1, 8.9, 5.9 Hz, 3H), 7.67 (d, J = 2.4 Hz, 1H),
7.55 (qd, J = 5.1, 1.8 Hz, 2H), 7.31 (dd, J = 8.9, 2.4 Hz, 1H),
3.02 (tt, J = 12.8, 3.4 Hz, 1H), 2.06 – 1.87 (m, 2H), 1.62 (t, J =
12.8 Hz, 2H), 1.14 (s, 6H), 1.11 (s, 6H). HRMS (ESI)+ m/z
calculated for C20H24NO3Na 349.1654 [M+Na]+ found
349.1658. FTIR (cm-1): 2760-3000 (rocking CH2), 1744 (C=O
ester), 1366 (N-O. stretch).
Synthesis procedure of 4HTs (4-hydroxyTEMPO derivatives):
To a stirred solution of 4-hydroxyTEMPO (1 eq.) DMAP (0.3
eq.) and the corresponding acid (0.7 eq) in dichloromethane
cooled to 0°C in an ice bath, was added EDC (1 eq), and the
mixture was warmed to ambient temperature over several hours
and stirred for 48-72 h, following the reaction advancement
through thin layer chromatography. The solution was diluted
with dichloromethane and was washed with water (2 ), a
saturated aqueous solution of sodium bicarbonate (1 ), and
brine (1 ). The organic layer was dried over sodium sulfate,
concentrated in vacuum and purified by silica gel column
chromatography to afford the compound as a reddish powder.
1
4HT-Benzene: (yield 53%) H NMR (300 MHz, DMSO-d6) δ
Experimental section
7.98 – 7.91 (m, 2H), 7.75 – 7.61 (m, 1H), 7.59 – 7.49 (m, 2H),
5.30 – 5.00 (m, 1H), 1.98 (ddd, J = 11.3, 4.3, 1.7 Hz, 2H), 1.61
(t, J = 11.7 Hz, 2H), 1.13 (s, 12H). HRMS (ESI)+ m/z calculated
for C16H22NO3 276.1600 [M]+ found 276.1603. FTIR (cm-1):
2834-3016 (rocking CH2), 1715 (C=O ester), 1363 (N-O.
stretch).
General procedure for the synthesis of TEMPO-bearing
derivatives (note that all characterizations relative to synthesis
of materials including, H NMR, FTIR, TGA/DSC and XRD
are reported at the end of the supplementary information file
(Figure S20 to S42) to facilitate the reading of the article):
1
1
4HT-Naphthalene: (yield 46%) H NMR (300 MHz, DMSO-
All commercially obtained solvents and reagents were used
without further purification unless noted below. 1-ethyl-3-(3-
d6) δ 8.67 – 8.55 (m, 1H), 8.18 – 8.09 (m, 1H), 8.08 – 7.91 (m,
3H), 7.74 – 7.58 (m, 2H), 5.26 (tt, J = 11.2, 4.2 Hz, 1H), 2.08 –
1.99 (m, 2H), 1.67 (t, J = 11.7 Hz, 2H), 1.15 (s, 12H). HRMS
(ESI)+ m/z calculated for C20H26NO3 328.1913 [M+2H]2+ found
328.1916. FTIR (cm-1): 2838-3016 (rocking CH2), 1712 (C=O
ester), 1365 (N-O. stretch).
dimethylaminopropyl)carbodiimide
dimethylaminopyridine (DMAP), 4-hydroxyTEMPO benzoate,
4-carboxyTEMPO, 4-hydroxyTEMPO, phenol, 2-
(EDC),
4-
hydroxynaphathalene, benzoic acid and 2-naphthalic acid were
purchased from Sigma Aldrich. The PTAm polymer has been
prepared and supplied by Nishide and Oyaizu’s group.
Chemical oxidation of 4CT-Naphthalene:
The chemical oxidation of 4CT-Napthtalene has been
performed following the already reported procedure16. Glacial
acetic acid (2 mL) and hydrogen peroxide (5 mL, 29.0–32.0%)
was added to 1 M NaClO4 aqueous solution (100 mL). The
solution was added dropwise to the stirred 4CT-Naphthalene
suspension in 1 M NaClO4 at ambient temperature. The mixture
was left under vigorous stirring during 24 hours.
Synthesis procedure of 4CTs (4-carboxyTEMPO derivatives):
To a stirred solution of 4-carboxyTEMPO (1 eq.) DMAP
(0.3 eq.) and the corresponding alcohol (0.7 eq) in
dichloromethane cooled to 0°C in an ice bath, was added EDC
(1 eq), and the mixture was warmed to ambient temperature
over several hours and stirred for 48-72 h, following the
reaction advancement through thin layer chromatography. The
2
ACS Paragon Plus Environment

Page 3 of 17
Chemistry of Materials
Electrochemical study:
Characterization techniques:
1H-NMR spectra were acquired using a Bruker ARX 300 MHz
spectrometer. Spectra were recorded at room temperature,
chemical shifts are written in ppm, and coupling constant in Hz.
Multiplicity is presented in the following way: s = singlet, d=
doublet, t = triplet, q = quintuplet, m = multiplet. High
resolution mass spectrometry (HRMS) has been recorded on a
XEVO G2-XS QTOF (Waters) operating on ESI+. Thermal
analyses were performed with a NETZSCH STA 449F3 device
under Argon atmosphere. Fourier transform infrared
spectroscopy (FTIR) spectra were collected with a Bruker
Vertex 70 device employing KBr pellets, using a DTGS
detector at a resolution of 4 cm-1. Powder XRD diagrams have
been recorded with the Bruker D8 Advanced system. X-ray
single crystal diffraction data of 4CT-Naphthalene were
collected on an Agilent SuperNova diffractometer equipped
with Atlas CCD detector and mirror monochromated
microfocus Cu-K radiation (λ = 1.54184 Å). The structure was
solved and refined on F2 by full-matrix least-squares techniques
using SHELX programs. All non-hydrogen atoms were refined
anisotropically and hydrogen atoms were placed in calculated
positions. Multiscan empirical absorption was corrected using
CrysAlisPro program (CrysAlisPro, Agilent Technologies,
V1.171.37.35 g, 2014). A summary of the crystallographic data
and the structure refinement parameters is provided in Table S1.
Operando X-ray diffraction patterns were recorded using a
PANanalytical X’Pert Pro diffractometer operated in Bragg-
Brentano reflection geometry with a Long Line Focus Cu-anode
X-ray source, and a X’Celerator RTMS detector. Data were
collected in a (10-30)° 2 range with a step of 0.033° and an
acquisition time of 5’16’’ per diagram. Measurements were
performed with a Multi Purpose Sample Stage (MPSS) able to
accommodate the in situ Swagelok-type cell. Note that the
common beryllium window was removed because it was
unstable in the tested electrochemical conditions. Therefore, we
replaced the beryllium window by a thin Kapton® foil (X-ray
film polyimide Kapton®, PANalytical B-V); thickness: 10 µm,
diameter: 6.4 mm. A threefold-size (in mAh) carbon electrode
(80 wt.% activated carbon Norit 1600, 10 wt.% Ketjenblack
The various electrochemical measurements were performed
using SP 300, MPG-2 and VMP3 potentiostats from Bio-logic
SAS (Seyssinet-Pariset, France). All the TEMPO derivatives
have been characterized by galvanostatic cycling in aqueous
NaClO4 1 M electrolyte. 4HT-Benzene was also tested in highly
concentrated aqueous electrolyte (8 M). Unless otherwise
specified, the electrode composition is 70 wt.% active material,
25 wt.% KB, 5 wt.% PTFE. The electrode mixtures have been
prepared vigorously mixing the three components in the dry
state with the use of a mortar. A co-precipitated mixture has
been also prepared. The 4CT-Naphthalene (70 wt.%) has been
dissolved in the minimum volume of dimethylsulfoxide
(DMSO) then 25 wt.% KB has been added to the solution. The
suspension has been vigorously stirred during one hour.
Subsequently, it has been added dropwise into a stirred water
solution. The precipitate has been finally filtered, washed and
dried at 60°C under vacuum overnight. 5 wt.% PTFE has been
finally added mixing the powders in a mortar. Caution was
taken in preparing thin electrodes with at 7 to 10 mg of active
material/cm² (0.7-1 mAh/cm²) pressed at 5 tons on a stainless
steel (AISI 316L) grid current collector. Thicker electrodes of
80 mg/cm² were made in two steps. First a pellet was obtained
by pressing at 5 tons the active material mixture. This pellet was
then introduced between two stainless steel grids and pressed at
2 tons in order to favor the current collector-electrode mixture
contacts. A threefold-oversized counter-electrode (80 wt.%
activated carbon Norit 1600, 10 wt.% KB, and 10 wt.% PTFE)
and a Saturated Calomel Electrode from ALS Co., Ltd Japan
(ref: RE-2BP) (SCE) as reference were used in all aqueous tests.
Thick electrode full cells have been prepared based on DNVBr
7.8 mAh/cm² (80 mg/cm², 1.09 mm for 4HT-benzene) and
8 mAh/cm² (80 mg/cm², 0.86 mm for DNVBr). Such cell
balancing was found to result in the highest energy density. All
full cells tests were performed within an argon-filled glovebox
with less than 1 ppm O2. An example of thick electrodes full
cell has been also made using a pouch cell three electrodes
configuration. Two microfiber glass separators were used, the
first to isolate the positive (81 mg/cm², 7.9 mAh/cm² for 4HT-
benzene) and the negative electrode (86 mg/cm², 9 mAh/cm²
for DNVBr) and another between the negative and reference
electrode. This cell balancing was chosen to maximize the
energy density. not screened for pouch cells. Pouch cells
reference electrodes were composed of Li0.5FePO4. The cell
used was an aluminum bag (8 cm  5.5 cm) from Sidec (model
CA 40) sealed at 190°C, where 1.5 mL of electrolyte (8 M
NaClO4) have been added. The pouch cell has been prepared in
an argon-filled glovebox and tested outside. Regular cyclic
voltammetry experiments in solution have been conducted by
dissolving 1 mM of the four TEMPO derivatives in
acetonitrile/NaClO4 2.5 M as the electrolyte and recorded with
a Pt micro disc electrode ( = 1.5 mm). A large Pt wire has
been used as counter electrode whereas the reference electrode
consisted in AgNO3/Ag from ALS Co., Ltd Japan (ref: RE-7,
TBAClO4 10 mM in acetonitrile). Potential values in Figure S1
are reported versus an aqueous SCE electrode (Ag/AgNO3 =
+0.36 V vs SCE). Temperature of the electrochemical tests is
22°C. Electrochemical impedance measurements were
performed from 180 kHz to 100 mHz with a perturbation
voltage of 10 mV. Power tests have been carried out on
discharge using the single discharge technique consisting of
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(KB)
EC-600JD
(AkzoNobel),
and
10 wt.%
polytetrafluoroethylene (PTFE)) was used as negative and
reference electrode whereas the positive electrode composition
was 70 wt.% of active material, 25 wt.% KB, and 5 wt.% PTFE.
This mixture was pressed at 10 tons on a stainless steel (AISI
316L) grid current collector. Thick electrodes of 4HT-benzene
and DNVBr were obtained pressing at 5 tons a 70 wt.% active
material, 25 wt.% KB, 5 wt.% PTFE mixture into a pellet. The latter
was then inserted in a specially designed mold and pressed at 2 tons
between two stainless steel grids. Peak fitting was performed
using Winplotr using a Gaussian function with constant FWHM
defined by U=0.07°. UV-Vis spectroelectrochemistry was
conducted by inserting a home-made three-electrode quartz cell
into a Shimadzu UV-2501 PC spectrometer. Each scan was
recorded from 250 to 800 nm within approx. 4 min. Calibration
curves were performed by dissolution of 4HT-Bz in acetonitrile
and 4CT-Napht in ethylacetate. Both these solvents show an
absorbance cut-off below 270 nm which allows to follow the
signal of the two compounds. Solutions of these compounds
show the same color as in the solid state although being more
pale as dissolution increases.
3
ACS Paragon Plus Environment

Chemistry of Materials
Page 4 of 17
relaxation periods until ΔE/Δt = 1 mV/h between each current Single crystals of the 4CT-Naphthalene compound could be
1
2
3
4
5
6
7
8
pulse from Imax to Imin. All experiments were conducted twice in
order to ensure reproducibility. Maximum experimental errors
related to capacity values were kept below 5% using our
protocols and set-up.
obtained and its crystal structure compared to that of one of the
two benzene derivative, the 4HT-Benzene. X-ray single crystal
structure data study reveals that 4CT-Naphthalene crystallizes
in the non-centrosymmetric P21 space group and consists of 1D
supramolecular arrangements of molecules (Figure 3a). The dN-
O distance of 1.286(3) Å found in 4CT-Naphthalene is typical of
a free radical single bond N−O17. Moreover, the chair
conformation of the piperidine ring clearly indicates that the
nitrogen is in sp3 hybridization state confirming N−O is a single
bond. The organization of 4CT-Naphthalene radicals into 1D
chains is achieved via weak C−H···O interactions between
oxygen and hydrogen of the radical rings with a N−O / O−N
distance of 5.766(2) Å (Table 1 and Figure 3b). In the bc plane,
N−O / O−N distances are shorter than in 1D chains (5.259(3)
Å). Comparing structural features of 4CT-Naphthalene with
that of 4HT-Benzene18 brings valuable information: N−O /
O−N distances in 4HT-Benzene are not only longer, but an
inversion also occurs as the shortest N−O / O−N distances run
along the 1D chains (6.128(7) Å) whereas the longest distances
are found in the bc plane (6.264(7)-6.287(6) Å).
Results and discussion
Four TEMPO bearing redox active molecules have been
investigated (Figure 2): the commercially available 4-hydroxy
TEMPO benzoate (4HT-Benzene), the isomer 4-carboxy
TEMPO benzoate and their naphthalene derivatives (4HT-
Naphthalene and 4CT-Naphthalene). The last three compounds
have been produced upon esterification reaction in order to
obtain the benzoate or naphthalate ester in the 4-position of the
piperidine ring. Importantly in regard to the targeted
application, these compounds can readily be obtained in gram
scales by a one pot reaction. The materials have been fully
characterized by means of Electrospray ionization high-
resolution mass spectrometry (ESI-HRMS), liquid nuclear
magnetic resonance (NMR) and Fourier transform infrared
(FTIR) spectroscopies as well as thermal analyses (for further
details see in supplementary information).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 1. Weak C-H···O interactions in 4CT-Naphthalene
C—H
(Å)
H···O
(Å)
C···O
(Å)
C—H···O
C—H···O (°)
i
C2—H2C···O1
0.96
0.98
2.62
2.64
3.553 (3)
3.623 (3)
163
178
i
C5—H5···O1
Symmetry code: (i) x+1, y, z.
Figure 2. Molecular structure of the four TEMPO derivatives
and their theoretical specific capacity.
4
ACS Paragon Plus Environment

Page 5 of 17
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
Figure 3. Structural features of 4CT-Naphthalene a) Crystal packing and N−O−O−N distance in the bc plane and b) N−O−O−N
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
distance into 1D chains and weak C-H···O interactions represented in dashed black line in 4CT-Naphthalene.
The electrochemical activity of the four materials has been
firstly characterized by cyclic voltammetry in the solubilized
state using acetonitrile/NaClO4 2.5 M as the electrolyte
(Figure S1). All derivatives exhibit a single reversible process
with Ep approaching the reversible one electron value of
58 mV. The characteristic peak potentials on oxidation was not
found to vary significantly indicating the substituent group has
a weak inductive effect if any. Lastly, derivation of diffusion
coefficients (Figure S2 and Table S2, see details in
supplementary information) are all similar and comparable to
the value reported for the 2,2,6,6-tetramethylpiperidinyl-N-oxyl
(TEMPO) radical13.
plateau type process (II). We note the capacitive contribution of
the carbon additive (25 wt.%) to the overall electrochemical
response is negligible (Figure S4). The two benzene derivatives
appear quite unstable in 1 M NaClO4 electrolyte showing an
important capacity fade during cycling (Figure 5). 4HT-
Benzene in particular shows a continuous fading of 70% in 100
cycles (123 hours). Accordingly, UV-Vis analysis of the
corresponding electrolyte upon 100 cycles clearly highlights the
dissolution of the material (Figure S5). Interestingly, the
introduction of more hydrophobic naphthalene moiety
significantly improves the cyclability at the expense of a lower
specific capacity. In particular, the initial capacity of 4CT-
Naphthalene isomer increases from 35 mAh/g up to 44 mAh/g
in 60 cycles followed by a capacity loss restricted to 4.5%
during the following 40 cycles. Consequently, no color change
could be detected in the electrolyte, while chemical oxidation
with hydrogen peroxide did not result in any apparent
dissolution of the compound (see Experimental part).
As shown by galvanostatic experiments, the electrochemical
profile is strongly modified in the solid state with the
occurrence of two electrochemical steps (Figure 4). This point
is reiterated using cyclic voltammetry for the two naphthalene
compounds (Figure S3). The reaction starts on oxidation by a
smooth increase of the incremental capacity (I) followed by a
5
ACS Paragon Plus Environment

Chemistry of Materials
Page 6 of 17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Galvanostatic curves for 4CT-Benzene (blue) on cycle 19, 4HT-Benzene (green) on cycle 1, 4CT-Naphthalene (pink)
on cycle 60 and 4HT-Naphthalene (violet) on cycle 22 at 0.075 A/g in 1 M NaClO4 aqueous electrolyte. Selected cycles
correspond for each compound to the maximum capacity obtained during the first 100 cycles as shown in Figure 5. The areal
capacity of electrodes are about 0.7 mAh/cm2.
Figure 5. Cyclability curves for the four compounds in 1 M NaClO4 at 0.075A/g (4CT-Benzene (blue), 4HT-Benzene (green),
4CT-Naphthalene (pink) and 4HT-Naphthalene (violet)). The areal capacity of electrodes are about 0.7 mAh/cm2.
To get a better insight, UV-Vis spectroelectrochemistry has
been performed during 158 hours while the sample was cycled
at 0.075 A/g (194 cycles) within the usual potential window of
(0; 0.75) V vs SCE. Simultaneously with the initial increase of
capacity during the first 60 hours, a UV band appears at 300 nm
(Figure 6). The latter rises during 80 hours and stabilizes
throughout the remaining cycles to 0.07mM. This value
corresponds in fact to a negligible part of the initial compound
(5 wt.%) (Figure S6). The question is now whether, this
concentration corresponds to a solubility limit or alternatively,
is limited because part of the active material does not participate
in the electrochemical reaction. A first hint is given by the fact
that irrespective of the electrolyte to active material mass ratio,
4CT-Naphtalene always gives the same order of maximum
specific capacity (40 ± 5 mAh/g, 52% of the theoretical
capacity). Besides, there is no clue so far regarding the source
of capacity limitation. In this goal the first issue to be addressed
consist in determining if, as suggested above, some grains are
poorly electrically connected, or alternatively if all grains react
but the electrochemical reaction is thermodynamically limited
6
ACS Paragon Plus Environment

Page 7 of 17
Chemistry of Materials
because of overwhelming charge repulsion and/or lack of space
to host ions within the oxidized host material.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Operando UV-Vis spectroelectrochemical spectra (right) and corresponding evolution of the discharge capacity of
4CT-Naphthalene upon cycling (left) measured at 0.075A/g in 1 M NaClO4 aqueous electrolyte and within (0; 0.75) V vs SCE
potential window. The areal capacity of electrodes are about 0.7 mAh/cm2.
To address this point, Operando XRD measurements of 4CT-
Naphtalene have been performed during galvanostatic cycling.
Noteworthy, each scan (10-30°) was 5’16’’ therefore enabling
a reliable description of the structural behavior by approx. 24
scans for each charge or discharge. The compound has been first
cycled within (-0.1; 0.95) V vs SCE. The corresponding overall
structural evolution is given in Figures 7a for the first cycle (the
first four cycles are shown in Figure S7). The initial phase (P0)
is characterized in Figure 7a by three main peaks at 12.26°,
13.61° and 14.62°. During the first part of the charge
(0<Q/Q0<15% or 0<Scan<10) the galvanostatic profile is
characterized by a smooth variation of the voltage that
corresponds to that observed for the potential in figure 4b
(0<Q<12 mAh/g). Surprisingly however, no angular shift was
detected for the initial set of peaks during the first 10th scans (55
minutes, Figure S8). Instead, intensities were found to decrease
at the expense of a new phase (P1) characterized by a set of
growing reflections at constant angle (15.22°, 15.77°, 19.49°
and 24.17°) (Figure 7a and S7). Peak fittings allowed a
quantitative description of the phase transformation progress
referred to as ξi ((Figure 7b, pink curve). The latter was derived
according to the ratio ξi = 1-[I(P0)i/I(P0)0] where I(P0)0 and
I(P0)i are the sum of the integrated intensity of the three main
peaks of (P0) listed above, at the beginning and, during the
electrochemical cycling respectively. The evolution of (P1)
(Figure 7b, violet curve) was tracked by the ratio between the
integrated intensity I(P1)i of the (P1) main peak at 15.22° on
cycle i, over that obtained by extrapolation of I(P1)i at 100% of
the theoretical capacity, referred to as I(P1)MAX (Figure S9).
Finally, the evolution of the specific capacity Qn normalized by
the theoretical one Qtheoretical is also shown in Figure 7b
(comparison of these curves vs the voltage is reported in Figure
S10). The trends of the structural and electrochemical
evolutions probed by ξi and Qn/Qtheoretical generally fits well for
the whole cycle and the similar slopes of 1-[I(P0)i/I(P0)0] and
I(P1)i/ I(P1)MAX) in shows (P0) and (P1) transforms into one
another during the reaction. However, at the beginning of the
charge and discharge the extent of the phase transformation is
up to 10% below that expected by referring to the relative
capacity Q/Q0. This type of “structural delay” has been observed
for the phase transformation between LiFePO4/FePO4 19–22 due to
macroscopic heterogeneity associated with the active material grain
connectivity to ionic and electronic percolating networks20. We
note that on the first cycle the capacity on charge is 28.8% higher
than on discharge while it is only 12.7% higher on the second cycle
(Figure S11). However, the amounts of P1 at the end of the 1st and
2nd charges are virtually unchanged (Figure S8) while a marked
inflection appears on first charge above 0.8V (Figure S12)
suggesting side reaction, most presumably oxygen evolution,
contributes to the coulombic inefficiency. Importantly, results
clearly show that a significant fraction of (P0) (48%) does not
react upon charge (Figure 7b). When the current is reversed,
88% of (P1) transforms back to (P0) which fits well with the
recovered specific capacity of 41 mAh/g. We note that the
residual trace at the end of the reduction (6%) of (P1) suggests
either kinetic limitation and/or oxidation of (P0) by dissolved
oxygen which was produced on charge.
In conclusion, based on a one electron reaction as proposed in
Figure 1, operando XRD shows that the specific capacity of
4CT-Naphthalene scales with the mass ratio of matter that is
involved in the electrochemical reaction. Accordingly, either
some grains are not connected to the electrical circuit and/or the
electrode highly suffers from kinetic hindrance causing some
grains to partially react. This last hypothesis can primarily stem
either from the efficiency of electronic percolation network
7
ACS Paragon Plus Environment

Chemistry of Materials
Page 8 of 17
within the composite electrode or, from the electrical
conductivity of the compound itself (electronic and/or ionic).
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. a) XRD diagrams collected for 4CT-Naphtalene during the first cycle within (-0.1; 0.95) V vs SCE showing the phase
transformation between the initial phase P0 and the new phase P1 characterized by a peak at 15.22°. b) Evolution of the
corresponding structural phase transformation derived by ξi = 1-[I(P0)i/I(P0)0], (P0) being the initial phase, compared to the
evolution of the fraction of the new phase (P1) and the relative specific capacity (Qn/Qtheoretical).
In this context, the 4CT-Naphthalene has been tested doubling
the amount of conductive carbon (40 wt.% of KB, 10 wt.% of
CNT) everything else being the same. Surprisingly enough, the
discharged capacity in 1M NaClO4 aqueous electrolyte is
virtually unchanged (38 mAh/g) (Figure 8). SEM images of this
sample are compared in Figure 9 to that containing 25 wt.% of
carbon additive. The two samples distinctively show the three
components of the composite electrodes. More particularly, the
two mixtures show 4CT-Naphthalene particles are nicely
wrapped by nano-grains (and nano-tubes in the case of the
50 wt.
additive does improve the performance up to a certain point
where the entire surface of the grains is covered (as observed
with 25 wt.% of carbon – hand mixed). It would therefore not
be surprising that decreasing the grain size while achieving a
kind of conducting coating would most likely greatly enhance
the electrochemical properties of this compound. In this regard,
Ketjenblack is presumably not the most appropriate carbon
additive. To nail our point, we provide in the supplementary
information the test of the well-known PTAm polymer using
the present electrode chemistry. Indeed, results reported in
Figure S15 show that the maximum specific capacity that could
be measured does not exceed 44 mAh/g (49% of the theoretical
capacity). This value significantly departs from that obtained
recently by Oyaizu and Nishide’s group using an optimized
electrode design with only 5 wt.% of single wall carbon
nanotubes and an optimized current collector7.
mixture) of the carbon additive which suggests the main cause
for the relatively low specific capacity of 4CT-Naphthalene is
not related to insulated grains but rather to kinetic hindrance.
The size of 4CT-Naphthalene grains and quality of the carbon
additive/active grain mixture could be modified by using an
alternative protocol to prepare the electrode. The latter consists
in co-precipitating a DMSO solution of 4CT-Naphthalene
containing 25 wt.% of carbon additive in water (see
Experimental part for more details). SEM images of the
corresponding electrode mixture show that carbon particles
leave this time some un-covered active material grains
(Figure 9). As a result, only 29 mAh/g could be attained with
this type of electrode which is approx. 25% lower than with the
two other electrode mixtures (Figure 8). Complementary
images taken at lower magnifications (Figure S13) reiterate
these observations and highlight that in all cases most 4CT-
Naphtalene grains are a few microns large. Accordingly, in an
attempt to unify these results, we believe the slow kinetics of
4CT-Naphthalene electrodes is primarily ascribed to the low
electrical conductivity of the compound. Its effect is all the
more detrimental to the performance that grains are generally
quite large. Therefore, the quality of the mixture with the carbon
8
ACS Paragon Plus Environment

Page 9 of 17
Chemistry of Materials
Figure 8. Galvanostatic curves of 4CT-Naphthalene mixed
1
2
3
4
5
6
with 25 wt.% KB (blue), 40 wt.% KB + 10% CNT (green) and
co-precipitated with 25 wt.% KB (red) obtained in 1 M
NaClO4 aqueous electrolyte (0.075 A/g). Inset shows the
specific capacity recovered on discharge after correction of
that associated with the carbon (Figure S14). The areal
capacity of electrodes are about 0.7 mAh/cm2.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
Figure 9. SEM images of 4CT-Naphthalene electrode mixtures with 50 wt.% (left) and 25 wt.% (middle) of conductive carbon
(KB). The co-precipitated electrode mixture with 25 wt.% of conductive carbon is also shown (right).
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As mentioned above, the oxidized form of the 4HT-Benzene
dissolves in 1 M NaClO4. Accordingly, a “water-in-salt”
approach was evaluated using a more concentrated NaClO4
electrolyte (8 M) without any other additives23. It seems
important to remind that the price of NaClO4, H2O is only 1-
2$/kg24. The material has been cycled at 0.075A/g in a (0;
0.9) V potential window vs SCE. Figure 10a,b highlights the
effect of the electrolyte concentration. First, an estimation of the
equilibrium potentials (defined as (Eeq  <E> = (Epa + Epc)/2)
shows the latter is 50 mV lower in the concentrated electrolyte
(8 M) than in the 1 M electrolyte solution which compares well
with the potential difference expected from the Nernst equation
(i.e., E = 53 mV). Second, the cyclability and Coulombic
efficiency of the material are significantly improved in the
concentrated electrolyte. Indeed, the initial capacity is nearly
the same than in the 1M electrolyte but remarkably, it increases
a bit from 52 mAh/g to nearly 60 mAh/g (62% of the theoretical
capacity) during the first 30 cycles and is then constant for the
next 70 cycles (123 hours in total). Notwithstanding, as shown
by the Ragone plot in Figure 10c only 64% of the theoretical
capacity could be achieved even at a C/10 rate. In conclusion,
despite a higher concentration of the electrolyte solves the
cyclability issue, the specific capacity of 4HT-Benzene is still
hindered.
To gain a better understanding of the 4HT-Benzene redox
behavior in concentrated NaClO4 electrolyte (8M),
electrochemical impedance spectroscopy (EIS) measurements
have been performed at different states of charge using a sample
containing 10 wt.% of carbon additive (Figure 11a,b). This
relatively low carbon content is used to exacerbate the electrical
response associated with the material itself. Except for the
spectrum corresponding to the initial state, all EIS spectra
measured during the electrochemical process show a semicircle
followed by a diffusion and capacitive part (Figures 11 and
S16). The significant change between the partially charged and
initial states indicates the resistance associated with the
semicircle can be ascribed to a charge transfer process.
Interestingly, results show the charge transfer resistance (36
Ohms) is almost stable all along the electrochemical reaction
suggesting that the oxidized phase shows similar electrical
properties to the reduced one. The decrease of the carbon
content to 10% does significantly reduce the capacity to only
11 mAh/g (82% lower than with 25 wt.%, Figure 11 c,d),
demonstrating a strong impact of the percolating network. EIS
results therefore leave us to believe that the specific capacity of
4HT-Benzene is mitigated by its poor electron conductivity
which would hamper the electron transport across the electrode
when particles are poorly connected to the carbon network. This
9
ACS Paragon Plus Environment

Chemistry of Materials
Page 10 of 17
hypothesis is supported by the trend of the charge transfer 10 wt.% of CNT hand mixed with the material (Figure 11d). As
1
2
3
4
resistance that scales almost linearly with the inverse of the
specific capacity (Figure 11c) and therefore with the number of
particles connected to the carbon network. Accordingly, the full
capacity (97 mAh/g) could be achieved with 40 wt.% of KB and
a result, both TEMPO derivatives (4HT-Benzene and 4CT-
Naphthalene) suffer from kinetic hindrance that inhibits the
performance.
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 10. a) Incremental capacity curve on cycle 4 derived from galvanostatic cycling of 4HT-Benzene at 0.075 A/g in highly
concentrated 8 M (blue) and 1 M (green) NaClO4; b) cyclability of 4HT-Benzene at 0.075 A/g in 8 M (blue) and 1 M (green)
NaClO4; c) Ragone plot of 4HT-Benzene in 8 M NaClO4. Electrodes contain 25 wt.% of carbon additive and the corresponding
areal capacity are about 0.7 mAh/cm2.
10
ACS Paragon Plus Environment

Page 11 of 17
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 11. a) EIS measurements performed at different states of charge using a sample of 4HT-Benzene containing 10 wt.% of
carbon additive in 8M NaClO4; b) Galvanostatic curve of the 4HT-Benzene (0.7 mAh/cm2) electrode containing 10 wt.% of
carbon additive in 8M NaClO4; c) Dependence of charge transfer resistance towards the specific capacity (the capacity has
been normalized to the theoretical capacity (Q0); inset of c) EIS measurements performed at the end of charge of samples of
4HT-Benzene containing 50 wt.%, 25 wt.% and 10 wt.% of conductive carbon; d) Ragone plot of 4HT-Benzene electrodes
containing 50 wt.%, 25 wt.% and 10 wt.% of conductive carbon.
To curtail the impact (mass and cost) of the concentrated
NaClO4 electrolyte at the cell level, a 4HT-Benzene thick
electrode of 7.8 mAh/cm² (80 mg/cm²; 1.10 mm) areal capacity
was designed and evaluated (Figure 12). During the first 68
cycles, the capacity sharply rises from 22 to 40 mAh/g which
most likely mirrors an activation process associated with the
wetting of the electrode porosity. The latter can be mitigated
however, by applying a floating period at the end of the charge
(0.9 V) until the current drops to 0.015 A/g (C/5) (Figures 12a
and S17). With that condition, the discharged capacity slightly
increases up to 66 mAh/g4HT-Benzene or 5.3 mAh/cm² (68% of the
theoretical capacity) after 535 cycles (38 days). The coulombic
efficiency also increases during the first 30 cycles to stabilize at
99.8%, suggesting an important contribution of the oxygen
evolution in the initial stages. A blank electrode containing only
the conductive additive indicates that half of the irreversibility
actually stems from side reactions going on at the surface of the
carbon (Figure S18). The latter sensibly attenuate with cycling
presumably due to the increasing acidity of the electrolyte from
neutral pH to pH=4 after 25 cycles (Figure 12b). Remarkably,
the power properties of the thick electrode are not strongly
impacted up to C-rate as it merely behaves like the thin ones
(Figure 12c). This therefore also serves to reiterate that
performance are limited by the efficiency of the electrode
chemistry in coping for the low conductivity of the material.
Results are therefore hardly comparable to those of Oyaizu and
Nishide’ group7 who demonstrated the only other thick (>3
mAh/cm2) organic based electrode. Indeed, in this case a high
utilization of the electrode can only be reached thanks to the use
of specific SWNTC conducting additive and 4-m-thick
conductive SWNT layer deposited at the surface of the
current collector.
11
ACS Paragon Plus Environment

Chemistry of Materials
Page 12 of 17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 12. a) Cyclability of 4HT-Benzene in 8 M NaClO4 aqueous electrolyte (80 mg/cm²); b) corresponding cyclability and
coulombic efficiency together with c) Ragone plots of both thin (10 mg/cm²) and thick electrodes (80 mg/cm²).
Although no further attempts have been done to improve the
overall kinetics of the electrodes, a full cell has been evaluated
with thick electrodes (mixed with 25 wt.% of KB) by coupling
the 4HT-Benzene positive electrode to a novel bipyridinium-
naphthalene diimide oligomer (referred to as DNVBr) as the
negative one. The formula and synthesis of DNVBr are
provided in the supplementary information. As demonstrated in
our previous work for a molecule of the same family25 and
illustrated in scheme S1, DNVBr is a p and n type active
material that releases anions and uptakes cation simultaneously
during its electrochemical reduction (charge of the full cell).
The present 4HT-Benzene/DNVBr full cell is therefore a
mixed, rocking-chair anionic12,26 cell (that corresponds on
charge to the transfer of anions from DNVBr to 4HT-Benzene)
and a dual-ion5 one simultaneously (with the transfer on charge
of cations from the electrolyte to DNVBr and corresponding
anions to the 4HT-Benzene). Although a detailed study of the
electrochemical performance of the DNVBr compound is out of
the scope of the present paper, its charge-discharge profile and
specific capacity retention are provided in Figure S19 for an 80
mg/cm2 electrode. The capacity ratio 4HT-Benzene:DNVBr
leading to the highest energy density was found to be 0.99 and
the areal capacity of both electrodes was approx. 8 mAh/cm²
(see Experimental part for more details regarding the electrode
preparation). For the sake of comparison, an identical full cell
but using thin electrodes of 0.7 mAh/cm² (10 mg/cm²
electrodes) has also been evaluated. The cell has been cycled at
0.075 A/g (1C) within a controlled voltage of (0; 1.8) V vs. SCE
(Figure 13a, b) by holding the end-of-charge potential until
0.015 A/g (C/5). Under these conditions, an average voltage of
0.96 V was obtained with a remarkably flat specific capacity of
30 mAh/gactive materials associated with an excellent coulombic
efficiency (99.6%) over 285 hours (12 days, 190 cycles)
(Figure 13b,c). These results compare well with that achieved
with the thin electrode full cell while allowing much higher
areal capacity of 4.6 mAh/cm² (vs 0.7 mAh/cm² for the thin
one). Remarkably such a high areal capacity competes well up
to 8C (Figure 13d). We note the slightly reduced discharge
voltage arises from a higher polarization associated with the
thick electrodes (Figure 13a). In addition is worth noting that
although a high maximum charge voltage of 1.8 V is applied,
self-pH-buffering of the electrolyte occurs to around pH=7, thus
mitigating current collector corrosion. The specific energy
density of this non-optimized thick electrode cell is 28 Wh/kg
per mass of both active materials at C-rate and 35 Wh/kg at C/8
rate, the latter rate being still acceptable for stationary storage.
This 4HT-Benzene//DNVBr cell is the first full organic cell
with close to millimeter thick electrodes to the best of our
knowledge. In comparison, the specific energy density obtained
for the thin electrode cell saturates to approx. the same value of
36 Wh/kg at C and C/2 rates. However, extrapolating these
results to a whole stack cell with electrodes of more than 700
cm2, shows that the lower fraction of inactive components
inherent to the use of millimeter thick electrodes enables
13.6 Wh/kgstack at 1C and 17.4 Wh/kgstack at C/8, which
represents gains of nearly 13% and 37% respectively, vs. the
same cell but having 0.7 mAh/cm² electrodes (12.7 Wh/kgstack
)
(Figure 13c, d) (see supplementary information for details
regarding calculations).
12
ACS Paragon Plus Environment

Page 13 of 17
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 13. a) Cell voltage (Ecell)-capacity profile of full cells of 0.7 (blue) and 8 mAh/cm² (red) electrodes at 0.075 A/g, b)
Cyclability of the full cells (violet thick electrodes full cell; blue thin electrodes full cell) at 0.075A/g along with Coulombic
efficiency; c, d) corresponding Ragone plots in 8M NaClO4.
In the light of these results, pouch cell configurations with
similar thick electrodes have also been evaluated to further
demonstrate the practicability of the present full organic
aqueous battery (see Experimental part for further details). The
capacity ratio 4HT-Benzene: DNVBr was set to 1.01. In these
conditions, an average output voltage of 0.78 V was obtained
with similar areal capacity (4.5 mAh/cm² corresponding to
30 mAh/gactive materials), high cyclability (98% of capacity
retention after 500 hours), coulombic efficiency (99.6%) at
0.075 A/g (Figure 14a) and rate capability (Figure 14b). More
interestingly, an even higher areal capacity (5 mAh/cm²) has
been further attained at 0.009 A/g (C/8).
13
ACS Paragon Plus Environment

Chemistry of Materials
Page 14 of 17
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 14. a) Cyclability of the full cells (violet pouch cell; red glass cell) along with coulombic efficiency in 8M NaClO4; b)
corresponding Ragone plots.
Conclusion
cycles in neutral and cheap 8 M NaClO4 electrolyte. Paired with
a new bipyridinium-naphthalene diimide oligomer a novel type
of aqueous battery that operates simultaneously in anion-
rocking chair and dual-cation/anion mode is demonstrated. The
full cell shows excellent cyclability and coulombic efficiency
offering 35 Wh/kgmaterials on discharge at C/8. We believe these
results represent an important advance towards cheaper and
more sustainable energy storage systems.
All the above results serve to demonstrate and rationalize the
potentialities and limitations of two small TEMPO molecules in
neutral molar aqueous electrolyte. For the first the
commercially available 4-hydroxy TEMPO benzoate molecule
has been proved to be highly efficient active material in
millimeter thick electrodes delivering up to 5.3 mAh/cm² at C
rate for over 500
14
ACS Paragon Plus Environment

Page 15 of 17
Chemistry of Materials
(7)
Hatakeyama-Sato, K.; Wakamatsu, H.; Katagiri,
R.; Oyaizu, K.; Nishide, H. An Ultrahigh Output
Rechargeable Electrode of a Hydrophilic
Radical Polymer/Nanocarbon Hybrid with an
Exceptionally Large Current Density beyond 1
A Cm −2. Adv. Mater. 2018, 30, 1800900.
Koshika, K.; Sano, N.; Oyaizu, K.; Nishide, H. An
Aqueous, Electrolyte-Type, Rechargeable
Device Utilizing a Hydrophilic Radical Polymer-
Cathode. Macromol. Chem. Phys. 2009, 210
(22), 1989–1995.
Koshika, K.; Sano, N.; Oyaizu, K.; Nishide, H. An
Ultrafast Chargeable Polymer Electrode Based
on the Combination of Nitroxide Radical and
Aqueous Electrolyte. Chem Commun 2009,
No. 7, 836–838.
ASSOCIATED CONTENT
Supporting Information. Further electrochemical details and all
the materials’ characterizations could be found in the Supporting
Information. This material is available free of charge via the
Internet at http://pubs.acs.org.
1
2
3
4
5
6
7
8
AUTHOR INFORMATION
Corresponding Author
* Fabrice Odobel Fabrice.Odobel@univ-nantes.fr
* Philippe Poizot philippe.poizot@cnrs-imn.fr
* Joël Gaubicher Joel.Gaubicher@cnrs-imn.fr
(8)
(9)
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Author Contributions
All authors have given approval to the final version of the
manuscript.
ACKNOWLEDGMENT
The authors would like to thank Dr. Olivier Crosnier (IMN) for his
constant technical assistance with operando XRD experiments, and
with Dr. Magali Allain from the University of Angers for her
important help regarding the single crystal structural resolution.
(10) Suga, T.; Konishi, H.; Nishide, H.
Photocrosslinked Nitroxide Polymer Cathode-
Active Materials for Application in an Organic-
Based Paper Battery. Chem. Commun. 2007,
No. 17, 1730–1732.
(11) Itoi, H.; Hasegawa, H.; Iwata, H.; Ohzawa, Y.
Non-Polymeric Hybridization of a TEMPO
Derivative with Activated Carbon for High-
Energy-Density Aqueous Electrochemical
Capacitor Electrodes. Sustain. Energy Fuels
2018, 2 (3), 558–565.
REFERENCES
(1)
Lithium Process Chemistry: Resources,
Extraction, Batteries, and Recycling; Chagnes,
A., Swiatowska, J., Eds.; Elsevier: Amsterdam
Boston Heidelberg, 2015.
(2)
Suo, L.; Borodin, O.; Sun, W.; Fan, X.; Yang, C.;
Wang, F.; Gao, T.; Ma, Z.; Schroeder, M.; vonꢀ
Cresce, A.; et al. Advanced High-Voltage
Aqueous Lithium-Ion Battery Enabled by
“Water-in-Bisalt” Electrolyte. Angew. Chem.
Int. Ed. 2016, 55 (25), 7136–7141.
(12) Koshika, K.; Chikushi, N.; Sano, N.; Oyaizu, K.;
Nishide, H. A TEMPO-Substituted
Polyacrylamide as a New Cathode Material:
An Organic Rechargeable Device Composed of
Polymer Electrodes and Aqueous Electrolyte.
Green Chem. 2010, 12 (9), 1573–1575.
(13) Suga, T.; Pu, Y.-J.; Oyaizu, K.; Nishide, H.
Electron-Transfer Kinetics of Nitroxide
Radicals as an Electrode-Active Material. Bull.
Chem. Soc. Jpn. 2004, 77 (12), 2203–2204.
(14) Itoi, H.; Hasegawa, H.; Iwata, H.; Ohzawa, Y.
Non-Polymeric Hybridization of a TEMPO
Derivative with Activated Carbon for High-
Energy-Density Aqueous Electrochemical
Capacitor Electrodes. Sustain. Energy Fuels
2018, 2 (3), 558–565.
(15) Ji, X.; Lee, K. T.; Nazar, L. F. A Highly Ordered
Nanostructured Carbon–Sulphur Cathode for
Lithium–Sulphur Batteries. Nat. Mater. 2009,
8 (6), 500–506.
(16) Lepage, D.; Michot, C.; Liang, G.; Gauthier, M.;
Schougaard, S. B. A Soft Chemistry Approach
to Coating of LiFePO4 with a Conducting
(3)
(4)
(5)
(6)
Yamada, Y.; Yamada, A. Superconcentrated
Electrolytes to Create New Interfacial
Chemistry in Non-Aqueous and Aqueous
Rechargeable Batteries. Chem. Lett. 2017, 46
(8), 1056–1064.
Wang, F.; Suo, L.; Liang, Y.; Yang, C.; Han, F.;
Gao, T.; Sun, W.; Wang, C. Spinel
LiNi0.5Mn1.5O4 Cathode for High-Energy
Aqueous Lithium-Ion Batteries. Adv. Energy
Mater. 2017, 7 (8), 1600922.
Dong, X.; Yu, H.; Ma, Y.; Bao, J. L.; Truhlar, D.
G.; Wang, Y.; Xia, Y. All-Organic Rechargeable
Battery with Reversibility Supported by
“Water-in-Salt” Electrolyte. Chem. - Eur. J.
2017, 23 (11), 2560–2565.
Chikushi, N.; Yamada, H.; Oyaizu, K.; Nishide,
H. TEMPO-Substituted Polyacrylamide for an
Aqueous Electrolyte-Typed and Organic-Based
Rechargeable Device. Sci. China Chem. 2012,
55 (5), 822–829.
15
ACS Paragon Plus Environment

Chemistry of Materials
Page 16 of 17
Polymer. Angew. Chem. Int. Ed Engl. 2011, 50
(30), 6884–6887.
Dependence of Phase Transition Pathways in
Storage Electrodes: Application to Nanoscale
Olivines. Chem. Mater. 2009, 21 (8), 1557–
1571.
1
2
3
4
5
6
7
8
(17) Pang, X.; Wang, H.; Ran Zhao, X.; Jun Jin, W.
Anionic 3D Cage Networks Self-Assembled by
Iodine and V-Shaped Pentaiodides Using
Dimeric Oxoammonium Cations Produced in
Situ as Templates. Dalton Trans. 2013, 42 (24),
8788–8795.
(23) Pasta, M.; Wessells, C. D.; Liu, N.; Nelson, J.;
McDowell, M. T.; Huggins, R. A.; Toney, M. F.;
Cui, Y. Full Open-Framework Batteries for
Stationary Energy Storage. Nat. Commun.
2014, 5, 3007.
9
(18) Pang, X.; Zhao, X. R.; Wang, H.; Sun, H.-L.; Jin,
W. J. Modulating Crystal Packing and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(24) Sodium Perchlorate,Sodium Perchlorate
Monohydrate,Blasting Industry-10 Year Gold
Supply Factory Supply - Buy Sodium
Magnetic Properties of Nitroxide Free Radicals
by Halogen Bonding. Cryst. Growth Des. 2013,
13 (8), 3739–3745.
Perchlorate Blasting Industry,Sodium
Perchlorate Monohydrate Blasting,Sodium
Perchlorate Anhydrous Perchloric Acid
Product on Alibaba.com
//www.alibaba.com/product-detail/Sodium-
Perchlorate-Sodium-Perchlorate-
(19) Leriche, J. B.; Hamelet, S.; Shu, J.; Morcrette,
M.; Masquelier, C.; Ouvrard, G.; Zerrouki, M.;
Soudan, P.; Belin, S.; Elkaïm, E.; et al. An
Electrochemical Cell for Operando Study of
Lithium Batteries Using Synchrotron
Radiation. J. Electrochem. Soc. 2010, 157 (5),
A606–A610.
(20) Ouvrard, G.; Zerrouki, M.; Soudan, P.; Lestriez,
B.; Masquelier, C.; Morcrette, M.; Hamelet, S.;
Belin, S.; Flank, A. M.; Baudelet, F.
Monohydrate-Blasting_60285442358.html
(accessed Jul 9, 2018).
(25) Perticarari, S.; Sayed-Ahmad-Baraza, Y.; Ewels,
C.; Moreau, P.; Guyomard, D.; Poizot, P.;
Odobel, F.; Gaubicher, J. Dual Anion–Cation
Reversible Insertion in a Bipyridinium–
Diamide Triad as the Negative Electrode for
Aqueous Batteries. Adv. Energy Mater. 2018,
8 (8), 1701988.
(26) Sano, N.; Tomita, W.; Hara, S.; Min, C.-M.; Lee,
J.-S.; Oyaizu, K.; Nishide, H. Polyviologen
Hydrogel with High-Rate Capability for Anodes
toward an Aqueous Electrolyte-Type and
Organic-Based Rechargeable Device. ACS Appl.
Mater. Interfaces 2013, 5 (4), 1355–1361.
Heterogeneous Behaviour of the Lithium
Battery Composite Electrode LiFePO4. J.
Power Sources 2013, 229, 16–21.
(21) Chang, H.-H.; Chang, C.-C.; Wu, H.-C.; Yang,
M.-H.; Sheu, H.-S.; Wu, N.-L. Study on
Dynamics of Structural Transformation during
Charge/Discharge of LiFePO4 Cathode.
Electrochem. Commun. 2008, 10 (2), 335–339.
(22) Tang, M.; Huang, H.-Y.; Meethong, N.; Kao, Y.-
H.; Carter, W. C.; Chiang, Y.-M. Model for the
Particle Size, Overpotential, and Strain
16
ACS Paragon Plus Environment

Page 17 of 17
Chemistry of Materials
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment