Transport and electron transfer kinetics of polyoxovanadate-alkoxide clusters

Article abstract of DOI:10.1149/2.1351902jes

A family of four polyoxovanadate-alkoxide (POV-alkoxide) clusters was prepared and electrochemical techniques were used to evaluate diffusion coefficients and electron transport across a range of oxidation states. Synthetic routes were developed to increase the alkyl chain length of the [V₆O₇(OR)₁₂] cores, increasing R from the previously reported R = CH₃, C₂H₅ to R = C₃H₇, C₄H₉. Whereas increasing chain length may enhance solubility, such modifications may also hinder diffusion and electron transfer by shielding the core, thus we experimentally determined these parameters using both cyclic voltammetry and rotating disk voltammetry. Increasing the alkyl chain length of the POV-alkoxide nanostructures from methoxide to butoxide changes the solubility from 0.205 to 0.297 M in acetonitrile. Although some variations in diffusion coefficients and heterogeneous electron transfer rate constants were observed across the suite of oxidation states from species to species, they range from 0.14 × 10^-5 cm²/s to 2.24 × 10^-5 cm²/s for D0 and 0.56 × 10^-3 cm/s to 209.00 × 10^-3 cm/s for khet. An increased chain length did not result in lower diffusion coefficients. Thus, we conclude that between C1 and C₄ chains, no shielding of the redox core occurs, nor is transport through solution systematically hindered.

Full text of DOI:10.1149/2.1351902jes

A464
Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
0
013-4651/2019/166(4)/A464/9/$38.00 © The Electrochemical Society
Transport and Electron Transfer Kinetics of
Polyoxovanadate-Alkoxide Clusters
1
,∗
2
2,3
Anjula M. Kosswattaarachchi, Lauren E. VanGelder, Olaf Nachtigall,
1
2
2,z
1,z
Joshua P. Hazelnis, William W. Brennessel, Ellen M. Matson, and Timothy R. Cook
1
Department of Chemistry, University at Buffalo, State University of New York, Buffalo, New York 14260, USA
Department of Chemistry, University of Rochester, Rochester, New York 14627, USA
Institut f u¨ r Chemie und Biochemie, Freie Universit a¨ t Berlin, 14195 Berlin, Germany
2
3
A family of four polyoxovanadate-alkoxide (POV-alkoxide) clusters was prepared and electrochemical techniques were used to
evaluate diffusion coefficients and electron transport across a range of oxidation states. Synthetic routes were developed to increase
the alkyl chain length of the [V6O7(OR)12] cores, increasing R from the previously reported R = CH3, C2H5 to R = C3H7, C4H9.
Whereas increasing chain length may enhance solubility, such modifications may also hinder diffusion and electron transfer by
shielding the core, thus we experimentally determined these parameters using both cyclic voltammetry and rotating disk voltammetry.
Increasing the alkyl chain length of the POV-alkoxide nanostructures from methoxide to butoxide changes the solubility from 0.205
to 0.297 M in acetonitrile. Although some variations in diffusion coefficients and heterogeneous electron transfer rate constants were
−
5
2
−5
2
observed across the suite of oxidation states from species to species, they range from 0.14 × 10 cm /s to 2.24 × 10 cm /s for D0
−
3
−3
and 0.56 × 10 cm/s to 209.00 × 10 cm/s for khet. An increased chain length did not result in lower diffusion coefficients. Thus,
we conclude that between C1 and C4 chains, no shielding of the redox core occurs, nor is transport through solution systematically
hindered.
©
2019 The Electrochemical Society. [DOI: 10.1149/2.1351902jes]
Manuscript submitted November 27, 2018; revised manuscript received January 24, 2019. Published February 8, 2019.
Redox flow batteries (RFBs) are an important electrochemical
versible redox reactions; (2) multiple reversible redox events over a
large electrochemical window; (3) high chemical, electrochemical,
and cell-component stability across all operative oxidation states; (4)
high solubility in organic solvents; and (5) composition from inex-
pensive, earth-abundant starting materials. Nanoscopic polynuclear
architectures are excellent candidates as high-energy capacity naRFB
active species, as they can be systematically modified to obtain multi-
electron transfer and long-term stability, while simultaneously ad-
energy storage (EES) technology due to their independent scalability
1
of energy and power. A myriad of redox-active molecules have been
explored as charge-carriers in RFBs and significant efforts have been
made to identify highly soluble charge-carriers that have long-term
stability, motivated by the terms contributing to the volumetric energy
density (Evol) of a RFB device:
1
23
E
vol
=
nVcell FCactive
[1]
dressing membrane crossover, solubility, and redox potential. How-
ever, there remains a poor understanding of approaches for the deriva-
tion of new classes of polynuclear assemblies with fundamental prop-
erties relevant for EES, particularly in nonaqueous media. Filling this
gap in knowledge requires comprehension of the structure-function
relationships that inform how size, nuclearity, ligand design, and
metal composition control the electrochemical and physicochemical
profiles of multimetallic charge carriers. Whereas pioneering work
on small organic molecules and coordination complexes reveals the
2
Where n is the number of electrons transferred, Vcell is the average
cell voltage, F is Faraday’s constant, and Cactive is the concentration of
the charge carrier. Non-aqueous RFBs (naRFBs) allow exploitation
of the properties of organic solvents, which include both wide redox
2
windows and operating temperatures. Although there are a number
of organic solvents that do not interfere with the charge/discharge
of active species, many are impractical for actual RFB devices. Be-
yond redox windows, cost, mass-transport properties, and membrane
24–28
power of leveraging molecular-level knowledge and design,
to
3
,4
compatibilities are all practical factors to be considered. Darling,
Gallagher, Brushett and co-workers have done a comprehensive cost
date such a systematic approach is underdeveloped for polynuclear
assemblies.
4
analysis for naRFBS and also studied transport characteristics in-
We have recently identified and characterized a series of discrete,
self-assembled polyoxovanadate-alkoxide (POV-alkoxide) clusters,
cluding area specific resistance (ASR), cross-over current density, and
coupling between crossover and capacity loss toward establishing a
23,29,30
[
V
6
O
7
(OR)12], as suitable charge-carriers for naRFBs.
initial work, it was demonstrated that the ligand modification at the
bridging alkoxide moieties from a -CH group to -C group results
in significant improvement in cluster stability during charge–discharge
In our
3
general quantification strategy for solvents and membrane separators.
DMF and DMSO were recognized as highly conducting solvents that
can be used in tandem with Nafion 211 to achieve an ASR target
3
2 5
H
2
23
of 2.3 ꢀ cm . Yan and co-workers identified acetonitrile as another
cycling. In addition, we note that the carbon number in the bridging
alkyl groups has significant bearing on the overall solubility of the
clusters. The results observed in our preliminary studies prompted the
evaluation of homoleptic ligand substitution of POV-alkoxide clus-
ters on the electrochemical and physicochemical behavior in solu-
tion, toward the ultimate goal of developing superior charge-carriers
for naRFBs. In this work, we explore the electrochemical conse-
quences of extending alkyl-chain length of the bridging alkoxide lig-
good solvent for naRFBs, due to its low viscosity and high ionic
5
conductivity.
6
–11
A large collection of coordination complexes,
organic
12–18
19–21
compounds,
and redox active polymers
have been explored
as components of naRFBs; however, the widespread commercializa-
tion of naRFB technologies remains limited due to high capital cost,
low power and energy density, and relatively rapid capacity fade over
long-term cycling—all factors that are critically linked to the physical
ands, with a focus on the comparative electronic transfer kinetics
1
,22
n
and electrochemical properties of charge-carriers.
To date, there
of [V
6
O
7
(OMe)12] (1), [V
6
O
7
(OEt)12] (2), [V
6
O
7
(O Pr)12] (3), and
n
remains a lack of suitable redox active molecules with the necessary
physicochemical properties to address the aforementioned challenges
facing improvements of existing naRFB technologies.
To enhance the viability of naRFB technologies, charge carriers
must be designed to feature (1) kinetically fast and chemically re-
6 7
[V O (O Bu)12] (4) POV-alkoxide clusters. Whereas increasing the
chain length is attractive from the standpoint of increasing solubility,
there have been no investigations on whether or not these changes
hinder electron transfer in a systematic way. Therefore, we have mea-
sured the diffusion coefficients (D
ligands and the various charge states of the ethoxide variant and the
heterogeneous electron transfer rate constants (k ) of all relevant re-
dox couples. We used acetonitrile as the solvent in this study as it
0
) of a suite of C1 to C4 alkoxide
∗
0
Electrochemical Society Student Member.
E-mail: trcook@buffalo.edu; ellen.matson@rochester.edu
z
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
A465
effectively solubilized the clusters at all-relevant oxidation states, en-
flushed with argon during the filtration. The first 10–20 mL of filtrated
solution were collected in a separate flask and filtrated again until no
fine V O was passing through the filter paper. The clear filtrate was
2 5
23
abling investigations of longer-term stability.
and k parameterize the mass transport and electron transfer
D
0
0
properties of an electrochemical reaction that have ramifications on
charge transfer resistance, and are therefore critically important for
establishing the viability of a charge carrier.31,32 The most common
collected in an argon flushed 1 L round bottom flask. After the fil-
tration, the solution was distilled under argon atmosphere. First, the
residual BuOH was distilled away from the crude product mixture at
n
methods used for analyzing RFB charge carriers are rotating disk
normal pressure. Subsequently, the temperature was increased and the
pressure was reduced to distil the vanadic acid ester from the crude
electrode voltammetry (RDE) and cyclic voltammetry (CV).8
,9,33
We
3
4
◦
report the values of D
0
obtained using the Randles-Sevcik method,
product. A cold trap (196 C) was installed between the distillation
35
36
◦
and estimated k values using the Nicholson, Kochi-Klinger, and
0
and the Schlenk line. Distillation (145 C, 1 mbar) gave the product
as yellow liquid (30.3 g, 14% based on vanadium). H NMR (CDCl
37
1
Koutecky-Levich methods. In addition to the electrochemical ki-
netics and transport parameters, we present the impact of bridging-
alkoxide ligand substitution on stability and solubility of the ho-
moleptic POV-alkoxide clusters in acetonitrile. Together, these rigor-
ous analyses establish chain-length independence of electron transfer,
suggesting ligand modifications as an ideal way to improve the en-
ergy density of POV-alkoxide derived charge-carriers for non-aqueous
electrochemical energy storage.
3
,
400 MHz) δ = 4.96 (br. s, 6H), 1.78 (m, 6H), 1.55 (m, 6H), 0.96 (t, J
1
3
= 7 Hz, 9H). C NMR (CDCl
3
, 100 MHz) δ = 82.4, 35.7, 19.0, 13.9.
51
V NMR (CDCl
3
, 105 MHz) δ = −595.
n
n
Preparation of [V
6
O
7
(O Pr)12] (3).—In the glove box, VO(O Pr)
3
n
(
4 4
0.300 g, 1.2 mmol), [ Bu N][BH ] (0.132 g, 0.5 mmol), and n-
propanol (10 mL) were charged in a 25 mL Teflon-lined autoclave.
The reactor was sealed, removed from the glove box and placed in an
oven where the mixture was heated to 125 C for 24 h. The autoclave
◦
Experimental
was removed from the oven and cooled to room temperature over three
hours. The subsequent workup was completed in ambient atmosphere.
The resulting deep red solution was left exposed to air for one hour,
during which the color slowly turned the distinct green of a mixed-
valent, hexavanadate cluster. The crude product was chromatographed
on a silica column (n-propanol), and the first, light green fraction
collected. The solvent was evaporated under reduced pressure, and
the product was extracted with acetonitrile, then dried under reduced
pressure to give the product, 3, as a sticky, green solid (0.125 g,
0.1 mmol, 54%). Crystals suitable for X-ray diffraction were grown
Synthesis and characterization.—General considerations.—Ma-
nipulations which required the absence of water and oxygen were
conducted in a UniLab MBraun inert atmosphere glove box un-
der a dinitrogen atmosphere. Glassware was oven dried for a min-
imum of 4 hours and cooled in an evacuated antechamber prior to
use in the glove box. Anhydrous methanol, ethanol, propanol, and
butanol were purchased from Sigma-Aldrich and stored over acti-
vated 4 Å molecular sieves purchased from Fisher Scientific. All
other solvents were dried and deoxygenated on a Glass Contour Sys-
1
from slow evaporation of n-propanol at room temperature. H NMR
tem (Pure Process Technology, LLC) and stored over activated 4 Å
−
1
n
(500 MHz, MeCN-d ): δ = 26.8 (br). FT-IR (ATR, cm ) 1036, 989
molecular sieves. V
2
O
5
(99.5% approx., 60 mesh), [ Bu
4
N][BH
4
],
3
−
1
−1
n
(O −R), 968 (V = O ). UV−Vis (CH CN) [ε (M cm )]: 384
3 3
and VO(O Pr)
received. [V
according to literature precedent.
3
were purchased from Sigma-Aldrich and used as
(OMe)12] (1) and [V (OEt)12] (2) were prepared
TBAPF was purchased from
b
t
3
nm (4.70 × 10 ), 1000 nm (1.24 × 10 ). ESI-MS(+): m/z 1127
6
O
7
6 7
O
n
23,38
[V O (O Pr) ]. Elemental analysis: Calculated (%) for (MW = 1127
6
7
12
6
g/mol): C, 38.38; H, 7.51%. Found: C, 38.40; H, 7.53%.
Sigma-Aldrich, recrystallized thrice using hot methanol, and stored
under dynamic vacuum for a minimum of two days prior to use.
1
n
n
H NMR spectra were recorded on a Bruker DPX-500 spectrometer
Preparation of [V O (O Bu) ] (4).—In the glove box, VO(O Bu)
6
7
12
3
n
locked on the signal of deuterated solvents. All chemical shifts were
4 4
(0.300 g, 1.0 mmol), [ Bu N][BH ] (0.090 g, 0.4 mmol), and n-
reported relative to residual proteo solvent resonances. CD
3
CN was
butanol (10 mL) were charged in a 25 mL Teflon-lined autoclave.
The reactor was sealed, removed from the glove box and placed in an
purchased from Cambridge Isotope Laboratories, degassed by three
freeze−pump−thaw cycles, and stored over activated 4 Å molecular
sieves. Infrared (FT-IR, ATR) spectra of complexes were recorded on a
Shimadzu IRAffinity-1 Fourier Transform Infrared Spectrophotome-
◦
oven where the mixture was heated to 125 C for 24 h. After this time,
the autoclave was removed from the oven and set to stand at room
temperature for three hours. The subsequent workup was completed
under an ambient atmosphere. The resulting deep red solution was left
exposed to air for one hour, during which time the color of the solution
slowly turned green. The crude product was then chromatographed
on a silica column (n-butanol), and the first, light green fraction was
collected. The solvent was evaporated under reduced pressure, and the
product was extracted with acetonitrile, then dried in vaccuo to give
the desired product, 4, as a dark green oil (0.099 g, 0.1 mmol, 44%).
–1
ter and are reported in wavenumbers (cm ). Electronic absorption
measurements were recorded at room temperature in anhydrous ace-
tonitrile in a sealed 1 cm quartz cuvette with an Agilent Cary 60 UV-
Vis spectrophotometer. Mass spectrometry analyses were performed
L
on an Advion expression Compact Mass Spectrometer equipped with
an electrospray probe and an ion-trap mass analyzer. Direct injection
analysis was employed in all cases with a sample solution in acetoni-
trile. Single crystals were mounted on the tip of a thin glass optical
fiber and positioned on a XtaLab Synergy-S Dualfelx diffractometer
equipped with a HyPix-6000He HPC area detector for data collection
1
−1
−
H NMR (500 MHz, MeCN-d ): δ = 26.6 (br). FT-IR (ATR, cm
3
)
1
1043, 1026, 1007 (O −R), 972 (V = O ). UV−Vis (CH CN) [ε (M
b
t
3
−
1
3
3
cm )]: 384 nm (5.93 × 10 ), 1000 nm (1.11 × 10 ). ESI-MS(+):
3
9
n
n
at 100.0(5) K. The structures were solved using SHELXT-2014/5
6 7 12 6 7 11
m/z 1295 (V O (O Bu) ), 1253 (V O (OMe)(O Bu) ). Calculated
40
and refined using SHELXL-2014/7. Elemental analyses were per-
formed on a PerkinElmer 2400 Series II Analyzer, at the CENTC
Elemental Analysis Facility, University of Rochester.
(%) for combined MW = 1285 g/mol ([V O (OCH ) (OC H ) ],
6
7
3
x
4
9
1−x
where x = 1 for 25% and x = 0 for 75%): C, 44.18; H, 8.36%. Found:
C, 44.178; H, 8.529%).
n
Preparation of VO(O Bu)
3
.—The preparation of this mononuclear
Solubility measurements.—The molar absorption coefficient of
each species was determined using five stock solutions, serially di-
luted to absorbances between 0.1 and 1.0 in accordance with the
Beer-Lambert relationship. Saturated solutions were prepared in trip-
licate by the sequential addition of solid into acetonitrile (3 mL) with
stirring until a suspension was formed. The suspensions were stirred
overnight, and then allowed to settle for at least 1 h, after which time
the solutions were filtered through glass wool to remove any undis-
solved material. An aliquot of each solution was diluted in acetonitrile
vanadium precursor is based on a modification of the previously pub-
lished procedure. A stirred suspension of V O (70.0 g, 385 mmol)
2 5
in 600 mL BuOH was heated in a 1 L round bottom flask equipped
with reflux condenser to 80–90 C under argon atmosphere overnight.
Stirring was then turned off and the temperature of the system was re-
was allowed to settle at the bottom of the flask for
0 min. The warm reaction mixture was gently decanted in portions
41
n
◦
◦
duced to 60 C. V
2
O
5
3
of 100–150 mL into a fritted filter funnel. The flask was constantly
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A466
Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
Figure 1. General synthetic scheme for the synthesis of POV-alkoxide
clusters.
and the UV-Vis spectra recorded to determine the maximum solubility
of each species.
Bulk electrolysis and stability studies.—Bulk electrolysis experi-
ments were performed with stirred solutions in an H-cell with a glass
frit separator (Porosity = 10–16 μm, Pine Research, USA) using
a Bio-Logic VMP3 potentiostat/galvanostat. An active species con-
centration of 1 mM was used. Each working electrode compartment
contained 15 mL of the active species solution with 0.1 M TBAPF
in CH CN, and each counter electrode compartment had 15 mL of
.1 M TBAPF in CH CN. A Pt mesh working electrode and a Pt
6
3
0
6
3
wire counter electrode were used. Bulk electrolysis experiments were
carried out using the chronoamperometry techniques available in the
Bio-Logic EC lab software suite at constant potentials, selected based
on CV experiments. All experiments were conducted at room temper-
ature inside a nitrogen-filled glove box (MBraun, USA).
Cyclic voltammetry (CV).—CV data were collected using a Bio-
Logic SP 300 potentiostat/galvanostat and the EC-Lab software suite.
Cyclic voltammograms were recorded using a 5.5 mm diameter plat-
inum fixed-disk PEEK shroud working electrode (PINE Research,
Durham, NC), a Pt wire auxiliary electrode (CH Instruments, USA),
+
and a Ag/Ag non-aqueous reference electrode with 0.01 M AgNO
3
6 3
in 0.1 M TBAPF in CH CN (BASi, USA). All measurements were
obtained at room temperature. Prior to data collection all solutions
were bubbled with nitrogen. Cyclic voltammograms were iR com-
pensated at 85% with impedance taken at 100 kHz using the ZIR
tool included within the EC-Lab software. All CV experiments were
repeated three times.
Figure 2. The NMR spectra (a), and the IR spectra (b) obtained for each
species.
reaction, the reaction mixtures of 3 and 4 are exposed to air, whereupon
they undergo a color change from red to green over the course of
thirty minutes. Once the green color is observed, characterization
by electrospray ionization mass spectrometry (ESI-MS) reveals m/z
peaks corresponding to those predicted for 3 and 4 (m/z = 1127 and
1295, respectively). Following purification via chromatography on
silica, complexes 3 and 4 are obtained in their neutral state, with
oxidation state distributions of vanadyl ions matching that of the
Rotating disk electrode (RDE) experiments.—A Pine MSR rota-
tor (PINE Research, Durham, NC) was used for RDE experiments.
The exact same configuration as the CV experiments were used to
avoid any changes due to physically different electrodes. The working
electrode was rotated at values spanning from 300 RPM to 2500 RPM,
while the voltage was linearly swept from 0 V vs. OCV to 1.5 V and
+
−
0.55 V vs. Ag/Ag non-aqueous reference electrode, at 5 mV/s for
IV
V
1
oxidation and reduction, respectively. All RDE experiments were re-
peated three times. The Koutecky-Levich analysis technique included
within the EC-Lab software were used to calculate the heterogeneous
electron transfer rate constants of each redox couple.
4 2
neutral clusters of 1 and 2 (V V ). Characterization via H NMR,
infrared, and electron absorption spectroscopies reveal the expected
features for a delocalized, mixed valent vanadium core (Figure 2). In
the case of 3, we were able to grow crystals suitable for X-ray analysis,
allowing for the unambigious confirmation of its structure (Figure 3).
Interestingly, although the ESI-MS of 3 shows only a single peak
Results and Discussion
n
corresponding to the fully propyl-bridged cluster, [V
6
O
7
(O Pr)12
]
Synthesis and characterization.—The general synthetic scheme
for the synthesis of POV-alkoxide clusters is shown in Figure 1. The
(m/z = 1127), in the spectrogram of 4, a second peak is observed
in addition to the expected peak of the fully butyl-bridged cluster
n
n
syntheses of the propyl and butyl clusters, [V
6
O
7
(O Pr)12] (3) and
[V
6
O
7
(O Bu)12] (m/z = 1295). This second peak, with m/z = 1253,
n
[
V
6
O
7
(O Bu)12] (4) were accomplished using protocols derived from
those established for [V (OEt)12] (2). To generate each of the longer
chain alkyl complexes, a vanadium(V) ester precursor (VO(OR) ; R
) was dissolved in the respective alcohol (ROH; R =
), and subjected to solvothermal conditions along with the
external reductant, tetrabutylammonium borohydride [ Bu
While in the case of the formation of the methoxide-substituted POV-
alkoxide cluster, [V (OMe)12] (1), the reducing nature of methanol
has a relative intensity of ∼30% that of the primary peak, and corre-
6
O
7
sponds to the substitution of a single bridging-butoxide for a bridging-
3
6 7 3 4 9
methoxide moiety, to generate [V O (OCH )(OC H )11]. Although
=
C
H
3
H
, C H
7
, C
4 9
4
H
9
clearly observable via ESI-MS, this “impurity”, appears to have no
impact on the redox profile of the cluster, and is not observed in the
C
3
7
n
1
4
N][BH
4
].
H NMR and IR spectra of 4. It is, however, detectable by elemental
analysis, which reveals carbon and hydrogen percentages that reflect
one quarter of the clusters having a single bridging-butoxide substi-
6
O
7
facilitates formation of the mixed-valent Linqvist ion, in the absence
6 7 3 x 4 9
tuted for a bridging-methoxide ([V O (OCH ) (OC H )1−x], where
of external reductant, the longer chain alkyl complexes 2, 3, and 4
x = 1 for 25% and x = 0 for 75%; calculated (%) for combined MW =
1285 g/mol: C, 44.18; H, 8.36%. Found: C, 44.178; H, 8.529%). At-
tempts to avoid this impurity via modifications to the synthetic routes,
n
all require [ Bu
4
N][BH
4
], in increasing proportions corresponding to
the increasing alkyl chain length. Following the 24 hour solvothermal
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
A467
Figure 4. CV profiles obtained for POV-alkoxide clusters at a Pt working
electrode using 1 mM of active species in 0.1 M TBAPF6 in CH3CN: scan
rate = 100 mV/s.
portant that no precipitation occurs during bulk electrolysis. We did
not observe any aggregation or precipitation during bulk electrolysis
experiments at the solubility limit of each neutral cluster, indicating
that the other charge states are equal or more soluble than this limiting
value. Thus, we took the concentration of the neutral cluster as the pa-
rameter for calculations of energy density. Using these solubility val-
ues we calculated the specific energy densities to be [V
6
O
7
(OMe)12] =
n
5
.14 kJ/L, [V
6
n
O
7
(OEt)12] = 8.38 kJ/L, [V
6 7
O
(O Pr)12] = 16.38 kJ/L,
and [V
6
O
7
(O Bu)12] = 52.44 kJ/L.
n
Figure 3. X-ray crystal structure of [V6O7(O Pr)12].
Bulk electrolysis and stability studies.—The CVs for all four POV-
alkoxide clusters are shown in Figure 4. The electrochemical stability
of complexes 1 and 2 were previously reported. The electrochemical
stability of complexes 3 and 4 was assessed using bulk electrolysis
or to separate it via chromatography, were unsuccessful. However,
owing to its apparent lack of impact on the electrochemical profile of
23
4
, we continued our analysis.
in conjunction with CV monitoring of the solutions. Bulk reduction
+
and oxidation of solutions of 3 and 4 at −1.1 and +1.0 V vs Ag/Ag ,
Solubility measurements.—The solubilities of complexes 3 and 4
were measured in acetonitrile in the presence of 0.1 M TBAPF , in or-
6
respectively, yielded no change in the redox profiles of the solution,
apart from the expected shift in open circuit potential (Figure 5). The
open circuit potential (OCP) of the propoxide cluster shifted from
der to mimic the conditions of a working, naRFB cell. The solubilities
were measured to be 0.097 M and 0.297 M for 3 and 4, respectively.
The limited solubility of 3 is contrary to our original hypothesis that
the increased alkyl chain length would promote solubility in organic
solvent. Instead, we find that while this propoxide-bridged cluster has
increased solubility over the ethoxide-bridged 2 (0.052 M), its solu-
bility is significantly less than that of the methoxide-bridged cluster, 1
+
+
the initial value of 0.12 V vs. Ag/Ag to 0.97 V vs. Ag/Ag upon
+
oxidation and −0.99 V vs. Ag/Ag upon reduction. For the butoxide
+
cluster the OCP value changed from 0.06 V vs. Ag/Ag to 0.93 V
+
+
vs. Ag/Ag and −1.07 V vs. Ag/Ag upon oxidation and
reduction, respectively. These OCPs are beyond the E1/2 values for
the corresponding redox events, as expected for successful electroly-
sis. The electrolyzed solutions were subsequently monitored for both
decomposition and self-discharge using CV and open circuit potential
measurements over the course of one week, confirming the stabil-
ity of both 3 and 4 in their charged states. The observed stability
of the propoxide- and butoxide-bridged polyoxovanadate clusters is
consistent with our previously noted trend that increasing the length
(
0.205 M). Interestingly, the butyl-bridged complex 4 has significantly
improved solubility over all other alkyl derivatives, and is notably the
only complex to be an oil at room temperature.
The least soluble component of an electrolyte couple in an RFB
determines the upper limit of concentration for calculations of energy
density. For species that cycle between multiple charge states, it is im-
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A468
Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
Figure 5. Bulk electrolysis and stability monitoring of complex 3 and 4 via cyclic voltammetry. Arrows indicate sweep direction and open circuit potential.
of the bridging alkyl chain in the POV-alkoxide clusters results in the
electrochemically stable complexes over all four redox events.
In Eqs. 2 and 3, n is the number of electrons transferred, α is the charge
23
2
transfer coefficient (α ∼ 0.5), A is the electrode area (0.2376 cm for
the Pt working electrode), c is the bulk concentration of the active
0
species, D is the diffusion coefficient of the active species and ν is
Transport and electron transfer kinetics of homoleptic POV-
alkoxide clusters.—Diffusion coefficients (D ) of the homoleptic POV-
alkoxide clusters.—The D values of all four POV-alkoxide clus-
0
0
ters were calculated using the Randles-Sevcik method. A scan-rate-
dependent CV study was performed, with POV-alkoxide clusters 1–4
in 0.1 M TBAPF in CH CN. Scan rates ranging from 25 mV/s to 1500
6 3
mV/s were used. The first oxidation (anodic) and first reduction (ca-
thodic) peaks of the neutral clusters were used to calculate the average
D
0
values. All redox couples in the study appeared quasi-reversible,
as we see both anodic and cathodic waves, yet the peak separations
8
,42,43
(
ꢁE
p
) were all > 59 mV.
Figure 6a shows a scan rate dependent
CV, obtained for the first oxidation and reduction redox couples of
complex 1. Similar CV profiles were obtained for 2–4. Linear relation-
ships were observed between peak currents and the square root of scan
rate for all redox couples in the study. In Figure 6b, peak current ver-
sus the square root of scan rate plots for the first anodic and cathodic
peaks of the methoxide cluster are shown. The observed linearity indi-
9
,10,44
cates diffusion controlled redox reactions.
For a reversible redox
couple, peak current, i is given by the following equation;
p
5
3/2
1/2ν1/2
i
p
= 2.69 × 10 n A cD
0
[2]
For an irreversible redox couple, the peak current, i
p
is given by the
following equation
Figure 6. Scan rate dependent CV profiles for the first oxidation and first
reduction redox couples of POV-methoxide cluster (a) and the peak current
versus square root of scan rate plots for redox couples of complex 01 (b).
= 2.99 × 10 n3/2α1/2 A cD
5
1/2ν1/2
[3]
i
p
0
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
A469
Table I. Diffusion coefficients of neutral POV-alkoxide clusters
calculated using the Randles-Sevcik method.
Reversible
Irreversible
Randles-Sevcik
equation (10 cm /s)
Randles-Sevcik
equation (10 cm /s)
−
5
2
−5
2
Species
V6O7(OMe)12
V6O7(OEt)12
0.17
0.15
0.15
1.39
0.18
0.24
0.25
2.25
n
V6O7(O Pr)12
n
V6O7(O Bu)12
Figure 8. Diffusion coefficients of POV-ethoxide cluster at varied redox states
calculated using Randles-Sevcik method.
became more negative (Figure 8), indicating that the anionic species
displayed better mass transport. The D
was the highest amongst all relevant charge states suggesting that the
hydrodynamic radius is smallest at this state.
0
of the +1 charged complex
Heterogeneous electron transfer rate constants (k
POV-alkoxide clusters.—The k values were determined using three
different methods, (1) the Nicholson, (2) the Kochi-Klinger, and
0
) of homoleptic
0
46
36
3
7
(
3) the Koutecky-Levich. The Nicholson and Kochi-Klinger meth-
Figure 7. Plots of diffusion coefficients calculated using the reversible and
irreversible Randles-Sevcik equations for the neutral POV-alkoxide clusters.
ods all use steady-state models for the calculation of electrode reaction
rates using CV data obtained at different scan rates and are based on
peak-to-peak separation (ꢁE
used to estimate k by relating ꢁE
parameter (ψ). The dimensionless parameter (ψ) is obtained by intro-
ducing the ꢁE values to the working curve described by Nicholson
p
). The Nicholson method is frequently
the scan rate. Since there is no equation to explicitly calculate the
diffusion coefficients for a quasi-reversible reaction, we report the
values obtained from the reversible and irreversible treatments for
0
p
with a dimensionless kinetic
p
upper and lower limits of D
0
. The estimated D
0
values obtained for
0
and Shain, and the heterogeneous electron transfer rate constant (k )
the neutral POV-alkoxide clusters using the Randles-Sevcik method
are shown in Table I.
for the redox reaction is calculated using the following equation:46
ꢀ
ꢁ
1
/2
0
πnF D ν
As shown in Figure 7, no significant trend was observed between
k
0
= ψ
[4]
D
0
and the length of the alkyl chain of the alkoxide ligands. Following
RT
n
the Stokes-Einstein relationship, we expect [V
6 7
O (O Bu)12] to have a
In Eq. 4, n is the number of electrons involved in the reaction, ν is the
scan rate, F is the Faraday constant, ψ is the Nicholson dimensionless
smaller D , as it is the largest amongst the four POV-alkoxide clusters
0
45
studied. Despite this expectation, the butoxide cluster demonstrated
higher values compared to other three clusters, indicative that between
C1 and C4 alkyl chains, the size of the alkoxide ligand is not a
critical determinant of diffusion behavior. Similar observations were
also made with chromium and vanadium acetylacetonate complexes,
parameter, D
T is the temperature. The Kochi-Klinger method also depends on
for estimating k0. The following equation is used to calculate
0
is the diffusion coefficient, R is the gas constant, and
ꢁE
p
the k
36
0
:
ꢂ
ꢃ
ꢄ
ꢅ
ꢀ
ꢁ
1
/2
2
where changing the ligand chain length resulted in only trivial changes
αnF D0ν
−α nF Epa − Epc
to the diffusion coefficients and electron transfer rate constants.25,26
k
0
= 2.18
exp
[5]
RT
RT
A similar correlation between D
observed for the ethoxide cluster. We calculated the D
chemically and electrochemically isolated redox states of the POV-
ethoxide. Both samples resulted in D values that are within the same
order of magnitude for +1, +2, −1, and −2 states as reported in
Table II. The D values increase as the overall charge of the clusters
0
values and charge state was
0
values of
Wherein α is the charge-transfer coefficient (∼0.5) and Epa and Epc
are the anodic and cathodic peak current potentials.
0
We also used RDE voltammetry as a third method for deter-
mining k
0
. RDE experiments were carried out over a range of ro-
0
tation rates from 300 rpm to 2500 rpm. Figure 9a shows the LSV
Table II. Diffusion coefficients calculated for all relevant redox states of the POV-ethoxide cluster.
−
6
2
−6
2
Reversible Randles-Sevcik equation (10 cm /s)
Irreversible Randles-Sevcik equation (10 cm /s)
Redox State
Electrochemically isolated species
1.5
Chemcially isolated species
1.7
Electrochemically isolated species
Chemcially isolated species
−2
−1
+1
+2
2.4
5.3
5.2
1.2
2.8
2.9
5.0
1.5
3.3
3.2
0.7
1.8
3.1
0.9
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A470
Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
Figure 9. RDE data for complex 01(a), Koutecky-Levich plots (b), and Tafel plots (c) for redox couples of POV-methoxide cluster.
profiles obtained in RDE experiments for 1. Similar LSV data
were observed with the other three alkoxide clusters as well. The
current responses at varied overpotentials were used to construct
Koutecky–Levich plots, which were used to determine mass-transfer-
microscopic models of electron transfer. The Marcus theory of elec-
tron transfer further states that the probability of electron transfer
events decreases with increasing distance between the donor and ac-
ceptor. It is therefore reasonable to expect changes in electron transfer
rate constants in POV-alkoxide clusters of varying side chain sizes.
The results show that regardless of the length of the alkyl chains
of the alkoxide ligands, these clusters all show fast electron transfer
kinetics as estimated from four different electroanalytical methods
and with no clear trends, i.e. the longer the chain, the slower the
electron transfer (Figure 10). We therefore conclude that the reorgani-
zation energy (λ) of POV-alkoxide clusters does not vary with the size
between C1 and C4 alkyl lengths, thus there is a minimal impact
on the overall electron transfer kinetics and transport phenomena.
This is a promising result given that a common strategy to in-
crease solubility is to modify pendant organic functionality and we
have also established that such modifications can increase stability,
thereby enabling multi-electron transfer. We expect to further ex-
plore the electrochemical and physicochemical behaviors of POV-
alkoxide clusters in other organic solvents and evaluate the per-
formance in varied membrane separators toward identifying more
compatible active species-membrane-solvent composites in the near
future.
k k
independent kinetic currents i for redox reactions. These i val-
ues were subsequently fit to the Butler-Volmer equation via Tafel
plots. Using the Tafel equation shown below, the rate constants were
calculated.
nFαη
log i
k
= log (nFcAk
0
) + 2
[6]
.303 RT
In Eq. 6, n is the number of electrons, α is the transfer coefficient, R is
the gas constant, T is temperature, k is the standard kinetic rate con-
0
stant, and η is the over potential. The Koutecky–Levich plots and
the corresponding Tafel plots for the four redox couples of methox-
ide cluster are shown in Figure 9. The average value of three tri-
als for the electron transfer rate constants are shown in Table III.
0
The estimated k values for all redox couples are in the range of
−
2
−3
1
0
cm/s - 10 7cm/s, values that are competitive with existing naRFB
4
charge-carriers.
The molecular structure of redox active species is expected to af-
fect electron transfer kinetics in both outer-sphere and inner-sphere
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
A471
Table III. Heterogeneous electron transfer rate constants from CV and RDE methods.
Heterogeneous electron transfer rate constant (cm/s)
CV
RDE
Species and Redox Couples
Koutecky-Levich (× 10−2)
Nicholson (× 10−3
)
Kochi (× 10−3
)
1
2
0
1
1
2
0
1
+/0
V6O7(OMe)12
POV
7.67 ± 0.62
7.30 ± 0.68
2.03 ± 0.03
0.86 ± 0.16
7.45 ± 0.09
5.06 ± 1.25
1.48 ± 0.33
0.68 ± 0.24
1.05 ± 0.01
1.39 ± 0.04
4.69 ± 0.01
0.24 ± 0.00
0.77 ± 0.11
2.94 ± 0.29
1.23 ± 0.08
20.00 ± 2.71
14.00 ± 6.44
0.93 ± 0.01
4.10 ± 0.15
3.50 ± 0.32
0.81 ± 0.02
3.10 ± 0.05
7.44 ± 0.63
7.39 ± 0.43
7.27 ± 0.43
2.92 ± 1.38
4.60 ± 0.08
7.60 ± 0.75
3.20 ± 0.04
0.52 ± 0.02
9.50 ± 0.97
11.60 ± 2.05
8.30 ± 0.08
+/1+
/1−
POV
POV
POV
POV
POV
POV
POV
POV
POV
POV
POV
POV
POV
POV
−/2−
+/0
17.00 ± 1.28
125.00 ± 12.45
209.00 ± 56.32
8.46 ± 6.24
V6O7(OEt)12
+/1+
/1−
−/2−
13.30 ± 1.27
25.00 ± 1.66
68.00 ± 1.53
21.00 ± 0.78
0.56 ± 0.06
n
1+/0
2
0
1
V6O7(O Pr)12
+/1+
/1−
−/2−
n
1+/0
2
V6O7(O Bu)12
52.00 ± 9.77
69.50 ± 9.85
13.70 ± 0.79
+/1+
/1−
0
Figure 10. Plots of heterogeneous electron transfer rate constants of each redox couple versus the chain length of POV-alkoxide clusters.
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

A472
Journal of The Electrochemical Society, 166 (4) A464-A472 (2019)
Summary
7. J. L. Barton, T. Carney, and F. R. Brushett, Meeting Abstracts, MA2016-02, 638
2016).
(
In this work, we evaluated electrochemical and physicochemical
properties of a series of POV-alkoxide clusters, toward developing
superior charge-carriers for naRFBs. We place an emphasis on eval-
uating whether ligand substitution has a deleterious effect on redox
8. Q. Liu, A. A. Shinkle, Y. Li, C. W. Monroe, L. T. Thompson, and
A. E. S. Sleightholme, Electrochemistry Communications, 12, 1634 (2010).
. Q. Liu, A. E. S. Sleightholme, A. A. Shinkle, Y. Li, and L. T. Thompson, Electro-
9
chemistry Communications, 11, 2312 (2009).
0. D. Zhang, H. Lan, and Y. Li, Journal of Power Sources, 217, 199 (2012).
1
behavior. The bridging alkoxide ligands in our POV clusters were
11. M. H. Chakrabarti, E. P. L. Roberts, C. Bae, and M. Saleem, Energy Conversion and
Management, 52, 2501 (2011).
12. A. M. Kosswattaarachchi, A. E. Friedman, and T. R. Cook, ChemSusChem, 9, 3317
n
modified to obtain [V
6
O
7
(OMe)12], [V
6
O
7
(OEt)12, [V
6
O
7
(O Pr)12],
n
6 7
and [V O (O Bu)12]. The butoxide cluster has the highest solubility
(2016).
of ∼0.3 M amongst the four clusters studied. At these cell voltages,
solubilities of ∼3 M are needed to match state-of-the-art energy den-
sities in all-vanadium aqueous RFBs. Alternative ligand substitutions
and heterometallic cores offer promising directions to improve both
cell voltage and solubility further, in an effort to raise energy densi-
ties even higher. The ethoxide, propoxide, and butoxide clusters all
demonstrate long term stabilities as shown by bulk electrolysis and
CV experiments. CV data enabled estimation of diffusion coefficients
by using peak current responses at different scan rates (the Randles-
Sevcik method). The heterogeneous electron transfer rate constants
were also determined in multiple ways: (a) the Nicholson and (b)
the Kochi-Klinger, using the peak separation potentials at varied scan
rates from CV experiments; and (c) the Koutecky-Levich analysis
using current extrapolated from RDE liner sweep voltammetry data
in the kinetically controlled region. These electrochemical studies re-
veal that the metal-alkoxide clusters uniformly exhibit fast electron
transfer kinetics in the regimes of 10 - 10 cm/s. In addition, we
observe no trend correlating the diffusion coefficient of a given POV-
alkoxide and the size of the bridging ligands. Due to the diffusion
controlled nature of all redox reactions involved and the fast electron
transfer kinetics, there will be minimal charge-transfer and mass trans-
port losses, regardless of alkyl chain length in EES devices that use
POV-alkoxide clusters as charge-carriers. Hence, it is concluded that
the improvements in solubility and long-term stability do not cost the
inherent fast electron transfer properties of the bridging ligand mod-
ified POV-alkoxide clusters. As a result, modifying the alkyl chains
of alkoxide ligands is an effective vector for tuning physicochemical
properties.
1
3. A. Iordache, V. Delhorbe, M. Bardet, L. Dubois, T. Gutel, and L. Picard, ACS Applied
Materials & Interfaces, 8, 22762 (2016).
14. J. Winsberg, C. Stolze, S. Muench, F. Liedl, M. D. Hager, and U. S. Schubert, ACS
Energy Letters, 1, 976 (2016).
1
5. C. S. Sevov, R. E. M. Brooner, E. Ch e´ nard, R. S. Assary, J. S. Moore,
J. Rodr ´ı guez-L o´ pez, and M. S. Sanford, Journal of the American Chemical Soci-
ety, 137, 14465 (2015).
16. S. Zhang, X. Li, and D. Chu, Electrochimica Acta, 190, 737 (2016).
1
7. X. Wei, W. Xu, J. Huang, L. Zhang, E. Walter, C. Lawrence , M. Vijayakumar,
W. A. Henderson, T. Liu, L. Cosimbescu, B. Li, V. Sprenkle, and W. Wang, Ange-
wandte Chemie International Edition, 54, 8684 (2015).
18. F. R. Brushett, J. T. Vaughey, and A. N. Jansen, Advanced Energy Materials, 2, 1390
(2012).
1
9. J. Winsberg, T. Hagemann, S. Muench, C. Friebe, B. H a¨ upler, T. Janoschka,
S. Morgenstern, M. D. Hager, and U. S. Schubert, Chemistry of Materials, 28, 3401
(2016).
20. G. Nagarjuna, J. Hui, K. J. Cheng, T. Lichtenstein, M. Shen, J. S. Moore, and
J. Rodr ´ı guez-L o´ pez, Journal of the American Chemical Society, 136, 16309 (2014).
1. S. H. Oh, C. W. Lee, D. H. Chun, J. D. Jeon, J. Shim, K. H. Shin, and J. H. Yang,
Journal of Materials Chemistry A, 2, 19994 (2014).
2. J. D. Milshtein, J. L. Barton, T. J. Carney, J. A. Kowalski, R. M. Darling, and
F. R. Brushett, Journal of the Electrochemical Society, 164, A2487 (2017).
2
−2
−3
2
23. L. E. VanGelder, A. M. Kosswattaarachchi, P. L. Forrestel, T. R. Cook, and
E. M. Matson, Chemical Science, 9, 1692 (2018).
4. A. P. Kaur, N. E. Holubowitch, S. Ergun, C. F. Elliott, and S. A. Odom, Energy
Technology, 3, 476 (2015).
2
25. P. J. Cabrera, X. Yang, J. A. Suttil, R. E. M. Brooner, L. T. Thompson, and
M. S. Sanford, Inorganic Chemistry, 54, 10214 (2015).
2
6. J. A. Suttil, J. F. Kucharyson, I. L. Escalante-Garcia, P. J. Cabrera, B. R. James,
R. F. Savinell, M. S. Sanford, and L. T. Thompson, Journal of Materials Chemistry
A, 3, 7929 (2015).
27. P. J. Cappillino, H. D. Pratt, N. S. Hudak, N. C. Tomson, T. M. Anderson, and
M. R. Anstey, Advanced Energy Materials, 4, 1300566 (2014).
8. J. Winsberg, T. Hagemann, T. Janoschka, M. D. Hager, and U. S. Schubert, Ange-
wandte Chemie International Edition, 56, 686 (2017).
2
2
3
3
3
3
9. L. E. VanGelder and Ellen M. Matson, Journal of Materials Chemistry A, 6, 13874
(2018).
Acknowledgments
0. L. E. VanGelder, B. E. Petel, O. Nachtigall, G. Martinez, W. W. Brennessel, and
E. M. Matson, ChemSusChem, 0.
1. S. Tejal and M. James, A Benchmarking Approach for Routine Determination of Flow
Battery Kinetics (2018).
2. J. D. Milshtein, K. M. Tenny, J. L. Barton, J. Drake, R. M. Darling, and F. R. Brushett,
Journal of The Electrochemical Society, 164, E3265 (2017).
3. A. Orita, M. G. Verde, M. Sakai, and Y. S. Meng, Nature Communications, 7, 13230
This research was supported by the University at Buffalo, SUNY
A.M.K., J.P.H., and T.R.C.) and the University of Rochester (L.E.V.
(
and E.M.M.). L.E.V. also acknowledges the National Science Foun-
dation for fellowship support (NSF-GRFP, DGE-1419118).
(
2016).
34. A. Sevcik, Coll. Czech. Chem. Comm, 13, 349 (1948).
5. R. S. Nicholson, Analytical Chemistry, 37, 1351 (1965).
36. R. J. Klingler and J. K. Kochi, The Journal of Physical Chemistry, 85, 1731 (1981).
ORCID
3
Timothy R. Cook https://orcid.org/0000-0002-7668-8089
3
3
3
4
7. J. K. A. V. G. Levich, Zhur. Fiz. Khim., 32, 1565 (1958).
8. C. Daniel and H. Hartl, Journal of the American Chemical Society, 127, 13978 (2005).
9. G. M. Sheldrick, SHELXT-2014/5.
References
0. G. M. Sheldrick, Acta Crystallographica Section C, 71, 3.
1
2
3
. B. Dunn, H. Kamath, and J.-M. Tarascon, Science, 334, 928 (2011).
. S.-H. Shin, S.-H. Yun, and S.-H. Moon, RSC Advances, 3, 9095 (2013).
. R. Darling, K. Gallagher, W. Xie, L. Su, and F. Brushett, Journal of The Electro-
chemical Society, 163, A5029 (2016).
41. W. Prandtl and L. Hess, Zeitschrift f u¨ r anorganische Chemie, 82, 103 (1913).
42. A. E. S. Sleightholme, A. A. Shinkle, Q. Liu, Y. Li, C. W. Monroe, and
L. T. Thompson, Journal of Power Sources, 196, 5742 (2011).
43. G. Denuault, M. Sosna, and K.-J. Williams, in Handbook of Electrochemistry,
4
5
6
. R. M. Darling, K. G. Gallagher, J. A. Kowalski, S. Ha, and F. R. Brushett, Energy &
C. G. Zoski Editor, p. 431, Elsevier, Amsterdam (2007).
Environmental Science, 7, 3459 (2014).
44. N. Aristov and A. Habekost, World Journal of Chemical Education, 3, 115 (2015).
45. M. Iwahashi and Y. Kasahara, Journal of Oleo Science, 56, 443 (2007).
46. R. S. Nicholson and I. Shain, Analytical Chemistry, 36, 706 (1964).
47. Y. Ding, Y. Zhao, Y. Li, J. B. Goodenough, and G. Yu, Energy & Environmental
Science, 10, 491 (2017).
. K. Gong, Q. Fang, S. Gu, S. F. Y. Li, and Y. Yan, Energy & Environmental Science,
8
, 3515 (2015).
. M. H. Chakrabarti, R. A. W. Dryfe, and E. P. L. Roberts, Electrochimica Acta, 52,
189 (2007).
2
Downloaded on 2019-02-12 to IP 129.81.226.78 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).