Designing high energy density flow batteries by tuning active-material thermodynamics

Article abstract of DOI:10.1039/d0ra10913d

The cost of electricity generated by wind and solar installations has become competitive with that generated by burning fossil fuels. While this paves the way for a carbon-neutral electrical grid, short- and long-term intermittency necessitates energy storage. Flow batteries are a promising technology to accommodate this need, with numerous advantages, including decoupled power and energy ratings, which imparts flexibility, thermal stability, and safety. Further, development of robust nonaqueous systems has the potential to greatly improve energy density, approaching that of lithium-ion batteries, while maintaining the advantages of flow systems. Herein we report a breakthrough on a bio-inspired nonaqueous redox flow battery (NRFB) electrolyte, which contains high-concentration active-material and maintains stability during deep cycling for extended time-periods. These advances are reinforced by thermodynamic considerations and computational investigations, which provide a clear path to further improvements. Electrochemical studies confirm that the active-material maintains its high stability at high concentration. This molecular scaffold clears two important hurdles in designing active-materials for nonaqueous electrolytes-low solubility and poor stability-providing an in-road to development of high-performance NRFB systems.

Full text of DOI:10.1039/d0ra10913d

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Designing high energy density flow batteries by
tuning active-material thermodynamics†
Cite this: RSC Adv., 2021, 11, 5432
a
a
Shyam K. Pahari,
Tugba Ceren Gokoglan,b Benjoe Rey B. Visayas,
Jennifer Woehl,a James A. Golen,a Rachael Howland,a Maricris L. Mayes,
a
b
a
*
Ertan Agar
and Patrick J. Cappillino
The cost of electricity generated by wind and solar installations has become competitive with that
generated by burning fossil fuels. While this paves the way for a carbon-neutral electrical grid, short- and
long-term intermittency necessitates energy storage. Flow batteries are a promising technology to
accommodate this need, with numerous advantages, including decoupled power and energy ratings,
which imparts flexibility, thermal stability, and safety. Further, development of robust nonaqueous
systems has the potential to greatly improve energy density, approaching that of lithium-ion batteries,
while maintaining the advantages of flow systems. Herein we report a breakthrough on a bio-inspired
nonaqueous redox flow battery (NRFB) electrolyte, which contains high-concentration active-material
and maintains stability during deep cycling for extended time-periods. These advances are reinforced by
thermodynamic considerations and computational investigations, which provide a clear path to further
improvements. Electrochemical studies confirm that the active-material maintains its high stability at
high concentration. This molecular scaffold clears two important hurdles in designing active-materials
for nonaqueous electrolytes – low solubility and poor stability – providing an in-road to development of
high-performance NRFB systems.
Received 29th December 2020
Accepted 12th January 2021
DOI: 10.1039/d0ra10913d
rsc.li/rsc-advances
while the availability of renewable energy is increasing, the load
on the electrical grid follows a distinct daily pattern that is
Introduction
decoupled from that of solar and wind output. Furthermore,
seasonal variability in the grid-load, as depicted in Fig. 1c,
highlights the need for longer duration storage, over months or
even years.2 As such, energy storage with sufficient capacity and
the exibility to level these short and long-term intermittencies
is a vital component of the nascent, renewable energy portfolio.
Among several technologies under development to meet the
storage needs of a modern, sustainable electrical grid, redox ow
Over the last decade, the cost of renewable energy has decreased
markedly. Since 2010, the cost of utility-scale energy from
photovoltaics has dropped more than 75% to less than $1/W
and the price of wind turbines has fallen 58%.1 In many areas
of the United States, the levelized cost of electricity from new
wind and solar farms is now lower than that from new natural-
gas-combined cycle power plants.2,3 Despite these rapid
improvements in the economics of renewable energy sources,
their intermittency on both short and long time-scales neces-
sitates energy storage. This can be readily illustrated with
comprehensive, current grid information made available by the
Energy Reliable Council of Texas (ERCOT), which manages
ꢀ90% of the electricity load in Texas.4 Notably, in that market,
wind-capacity alone occasionally exceeds 50% of the total grid
load.4 On top of this substantial wind capacity, by 2022 ERCOT
projects that solar capacity will have increased seven-fold
compared to 2019 (see Fig. 1a).4 As shown in Fig. 1b, however,
batteries (RFBs) have received considerable recent interest.5
A
principal advantage of these systems is their uncoupled power and
energy ratings, providing them excellent exibility for a range of
stationary storage applications.6 Seasonal load variations (such as
those apparent in Fig. 1c) could be accommodated by high-
capacity installations, while simultaneously mitigating short-
term, relatively higher power storage requirements (such as
those apparent in Fig. 1b). Development of such systems, with
capacity for long-duration storage, that can also provide storage for
daily cycling, has been identied as an important milestone in
achieving zero-carbon electricity goals.2
While the majority of RFB systems studied thus far have
been based on aqueous electrolytes,5 the modest electro-
chemical window in which water is thermodynamically stable to
electrolysis has led researchers to explore nonaqueous electro-
lytes.7 By implementing aprotic solvents, nonaqueous redox
aDepartment of Chemistry and Biochemistry, University of Massachusetts Dartmouth,
MA 02747-2300, USA. E-mail: pcappillino@umassd.edu
bDepartment of Mechanical Engineering, Energy Engineering Graduate Program,
University of Massachusetts Lowell, Lowell, MA 01854, USA
† Electronic supplementary information (ESI) available. CCDC 2045253–2045256.
For ESI and crystallographic data in CIF or other electronic format see DOI:
10.1039/d0ra10913d
5432 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
Fig. 1 New (bar) and cumulative (line) solar capacity reported by ERCOT (a), normalized grid load and renewable capacity in Texas for July 15–21,
2019 (load and wind) and July 13–19, 2020 (solar) (b), and weekly average normalized grid load and wind output in Texas during 2019 (shaded
region highlights the increased grid load during summer months) (c). Data provided by the Energy Reliability Council of Texas in ref. 4.
ow batteries (NRFBs) expand the thermodynamic potential their solubility, which is recognized as a critical parameter in
window at which a ow battery may operate.8 Notwithstanding determining energy density,32,35 and demonstrates their poten-
kinetic stabilization of aqueous systems,9 which allow for tial as a scaffold for development of next-generation ow battery
metastable electrolytes, if an appropriate NRFB chemistry can electrolytes. These gains are supported by electronic structure
be elucidated the energy density of ow batteries could be calculations and placed in a theoretical framework that
greatly improved. The need for robust, high energy-density considers active material solubility (SVBH), resulting from the
active materials for NRFB electrolytes has led to investigation interplay between two thermodynamic quantities: the free-
of an array of organic7,10–21 and metalorganic7,9,14,22–29 redox energy of active-material/solvent interactions (DGs*ol, Fig. 2)
couples, as well as redox polymers.7,14,30,31 While a recent techno- and free-energy of the active-material crystal lattice (DG*lat
,
economic analysis concludes that a pathway exists to cost- Fig. 2).35–37 This framework provides a design-strategy for
effective energy storage based on NRFB, thus far, poor chem- further gains in VBH solubility and lays a blueprint for tuning
ical stability, leading to low cyclability and low solubility are key
remaining research challenges that have prevented imple-
mentation of this technology.32
Recently, a catholyte was reported that meets the challenging
chemical-stability requirements of an NRFB system by using
a molecule that is produced biologically, rather than one
elucidated by human-design.33 Biosynthesis of the vanadium
compound known as Amavadin (see Fig. S1†) evolved naturally
in mushrooms from the Amanita genus, under selection pres-
sure for strong and specic vanadium binding, resulting in the
most stable vanadium species that has been reported to date.33
The analogous, dianionic coordination compound, vana-
dium(4+) bis-hydroxyiminodiacetic acid ([VBH]) can be synthe-
sized at large scale, using inexpensive reagents and exhibits
high chemical stability, even under cycling at high current and
to deep states-of-charge.34 This stability provides a robust
system with which to explore the deep parameter space of
solvents, salts and cell components that make NRFBs such an
Fig. 2 Schematic representation of the thermodynamic cycle for
attractive frontier for energy storage research.
Herein we report a breakthrough in the design and synthesis
of VBH-based NRFB active-materials that greatly improves on
dissolution of [cation][VBH] complexes. The free-energy of dissolution
ðDG*disÞ determines the solubility of the complex and results from the
interplay between the free-energy of the lattice ðDG*latÞ and that of
stabilizing solute–solvent interactions in solution ðDG*solÞ.
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5433

View Article Online
RSC Advances
Paper
Fig. 3 Scheme for preparation of arbitrary cations of [VBH] by metathesis of the calcium ion. NXXXX, X ¼ 1, 2 3 and 4 correspond to the tet-
ramethylammonium, tetraethylammonium, tetrapropylammonium, and tetrabutylammonium cations, respectively.
VBH concentration in various electrolyte compositions. cation interactions in the crystalline solid. The thermody-
Crucially, electrochemical studies, also reported herein, namics of active-material dissolution are governed by the
demonstrate that VBH maintains high stability in concentrated following equation:35–37
À
Á
electrolytes during cycling to deep state-of-charge, over an
extended duration.
DGd*is ¼ DGl*at þ DGs*ol ¼ ꢁRT ln SV* BH ꢂ Vm
(1)
In this equation, S*VBH refers to the molar solubility of
a [cation][VBH] complex, DG*dis refers to the free-energy of
dissolution of the complex, DG*lat refers to the lattice free-energy
of the complex, DG*sol refers to the free-energy of solute–solvent
interaction, and Vm is the molar volume of solid [cation][VBH].
To systematically probe these quantities, a strategy to replace
calcium (2+) with arbitrary cations was devised, in which CaVBH
is treated with the uoride salt of the targeted cations (see
Fig. 3). Aer ltration to remove the insoluble uorite salt, this
Results and discussion
[VBH] cation metathesis
While the calcium(2+) salt of [VBH] (CaVBH) can be conve-
niently prepared at large scale, with high yield, this compound
exhibits low solubility in all solvents investigated.33,34 We
reasoned that replacing the calcium ion with alkylammonium
cations would signicantly alter solubility ðS*VBHÞ by affecting
solute–solvent interactions in the electrolyte as well as anion–
Fig. 4 X-Ray crystallographic data, and mass spectrometric characterization of [NXXXX]1,2[VBH]. (a) and (b) correspond to thermal ellipsoid plots
of N2222 and N4444, bis-alkylammonium, vanadium(4+) [VBH], respectively and (c) and (d) correspond to those of N3333 and N4444 mono-alky-
lammonium, vanadium(5+) [VBH], respectively. Structures of key ions present in the mass spectrometry are provided in (e) and (f) corresponds to
the positive and negative mode mass spectra of [NXXXX]2[VBH], X ¼ 1 to 4. Further details of X-ray crystallographic measurements can be found in
ESI.†
5434 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
solution can be used directly as a battery electrolyte or the compared to the protons on the opposite face (H1 and H2,
[cation][VBH] salt can be isolated as a solid. This strategy was Fig. 5) as a result of shielding from a more electron-rich quad-
used to prepare a series of symmetric, bis-alkylammonium salts rant of the coordination sphere. These four ligand proton
of [VBH], having from one to four methylene units in the alkyl positions are related to another set by a C2 operation through
groups adjacent to the central nitrogen atom (NXXXX, where X ¼ the vanadium center (see Fig. 5). This description is roughly
1–4 methylene units (see Fig. 3). Removal of solvent affords consistent with the assignment reported for the natural product
crystals of alkylammonium [VBH] compounds suitable for X-ray Amavadin,40 which contains two chiral methine positions in
crystallography (see Fig. 4a, and b). Further, chemical oxidation each ligand, in place of the methylene positions in [VBH].
of the vanadium(4+) species, [N3333]2[VBH] and [N4444]2[VBH],
allows preparation of the oxidized, vanadium(5+) species, as
Solubility of alkylammonium [VBH] compounds
characterized by X-ray crystallography (see Fig. 4c and d,
Metathesis of the calcium(2+) ion in the manner described
respectively). Mass spectrometry indicates the presence of the
above and depicted in Fig. 3 profoundly affects the solubility of
parent [VBH] ion in solution as well as monocation adducts (see
the [cation][VBH] active-materials. As shown in Table 1, CaVBH
Fig. 4e and f).
is insoluble in all solvents tested except for water and DMSO, in
In addition to the X-ray crystallography and mass spec-
which it is modestly soluble. A clear trend emerges, in which
trometry discussed above, 1H NMR spectroscopy provides
VBH exhibits increasing solubility with increasing alkyl-chain-
insight into the structure of the [VBH] active-materials. Inte-
length. It is also notable that the oxidation state of the cation
1
gration of the H spectra of [N3333][VBH] and [N4444][VBH] (see
plays an important role in solubility, with the reduced, dianionic
Fig. 5) exhibits ratios of [VBH]-methylene protons to alky-
species exhibiting greater solubility (vide infra). Formulation of
lammonium–methylene and –methyl protons of ꢀ8 : 8 : 8 : 12
NRFB electrolytes containing active-material and support cation
and 8 : 8 : 8 : 8 : 12, respectively, conrming the expected stoi-
with concentrations approaching one and two mol Lꢁ1, respec-
chiometries of the ions. In both compounds, the methylene
tively, constitutes a signicant milestone. Careful consideration of
region of the spectra, between ꢀ4.3 and 5 ppm contains eight
these solubility trends and casting them in the context of a more
peaks which can be assigned to four sets of equivalent protons
general theoretical framework, has the potential to lead to further
of the [VBH] ligands, each split by a strongly coupled, geminal
improvements and provides general insight into the development
proton (2JHH ¼ 16 Hz).38,39 DFT-predicted NMR shis (see Fig. 5,
of high-energy density NFRB electrolytes.
inset) aid in assigning these peaks, with the two protons in the
same face as the h2-NO ligand (H3 and H4, Fig. 5) up-eld
Thermodynamic considerations for [cation][VBH] solubility
Solubility trend in the steric bulk of the [VBH]2ꢁ counter-
cation(s). To better understand the solubility trends discussed
above, DFT calculations were performed on the alkylammo-
nium cations and the VBH anions using their crystal structures
as input geometries.35 As noted by Thompson et al.,35 calcula-
tion of solvation energy DG*sol (see Fig. 2), or the difference
between gas-phase and solvated ions, is straightforward and the
values for each of the [NXXXX]y[VBH] species (y ¼ 1 or 2 for
oxidized and reduced active-material, respectively) for which
solubility data has been measured is listed in Table 2. These
calculations were carried out using a model known as SMD,41 in
which the solvent is considered a continuous dielectric
medium42 interacting with the solute electron density. This
model accounts for electrostatic interactions of a polarizable
continuum stabilizing the permanent charges on the cation and
anions, as well as short-range dispersion interactions between
the solvent and solute molecules in the rst solvation shell (see
Fig. 6a).41 Interestingly, the trend in DGs*ol runs counter to the
experimentally observed solubility data. Electrostatic potential
plots, shown in Fig. 6, prepared using natural charges output
from natural bond orbital analysis,43 provide an explanation for
this effect based on the distribution of positive charge on each
cation. In all four cases (NXXXX, X ¼ 1–4) the central nitrogen and
carbon atoms are relatively electroneutral while the bulk of the
positive charge is concentrated on the protons bonded to the
Fig. 5 NMR spectra of the vanadium(5+) complexes of [N3333][VBH]
(top, red) and [N4444][VBH] (bottom, blue). Insets show the stick
carbon directly adjacent to the nitrogen atom. This effect is
readily apparent in the 1H-NMR spectroscopy of the alky-
diagram and predicted NMR chemical shifts of the [VBH] anion, as well
as the stick diagrams of the [N3333] and [N4444] cations.
lammonium ions (see Fig. 5), with a large downeld shi of the
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5435

View Article Online
RSC Advances
Paper
Table 1 Solubility of [NXXXX]y[VBH] active-materials in various solvents, X ¼ 1 to 4, y ¼ 1 for vanadium(5+) species and y ¼ 2 for vanadium(4+)
species. Values correspond to the maximum concentration of [VBH] in the electrolyte in mol Lꢁ1, as determined by UV-Vis spectroscopy. Values
of dielectric constant taken from ref. 42
Vanadium(4+) species
Vanadium(5+) species
Dielectric
constant
+
+
+
+
+
+
Solvent
[Ca]2+
[N1111 2
]
[N2222 2
]
[N3333 2
]
[N4444 2
]
[N3333
]
[N4444]
Tetrahydrofuran
Acetonitrile
Dimethylsulfoxide
Propylene carbonate
7.5
37
47
66
Insoluble
Insoluble
0.54
Insoluble
0.004
0.11
Insoluble
0.33
0.53
Insoluble
1.09
0.83
0.17
0.80
0.69
0.77
Insoluble
0.12
0.62
0.008
0.51
0.55
0.65
Insoluble
0.03
0.28
0.68
0.27
methylene protons adjacent to the nitrogen atom. These elec- alkyl-chain length must arise from destabilization of the crystal
tropositive protons become less solvent-accessible as the alkyl- lattice. This is in accord with reported values of lattice energy
chain lengthens, preventing stabilizing dipole interactions for alkylammonium halides derived from thermochemical
between the solute and the solvent continuum and resulting in measurements, which show less negative lattice energy with
a less negative DGs*ol
.
longer alkyl-chain substituents.44–46 This observation also reso-
Considering the trend in solvation energies of the alky- nates with the observed high solubility of so-called non-
lammonium species discussed above, it is clear that any coordinating anions, which is explained by low lattice energy
meaningful theoretical prediction of active-material solubility inherent to bulky anions.47 Shown schematically in Fig. 7, with
must account for the lattice energy of the solid, at least quali- increasing alkyl-chain length, the effect of lattice destabilization
tatively. Furthermore, the experimentally observed trend of ðDG*latÞ overcomes the destabilization of the solvated ions
steeply increasing solubility for [NXXXX]y[VBH] with increasing ðDG*solÞ, resulting in an overall more negative DGd*is, thus
a higher solubility (S*VBH, eqn (1)) for the bulkier cations.
Notably, the solubility of CaVBH in DMSO, which is higher than
that of [N1111]2[VBH] and comparable to [N2222]2[VBH], is an
outlier in the trends discussed above. This may be the result of
Table 2 DFT-calculated solvation energies ðDG*solÞ for reduced and
oxidized [NXXXX]y[VBH] compounds in acetonitrile, X ¼ 1 to 4, and y ¼ 2
for reduced species and y ¼ 1 for oxidized species
specic solute–solvent interactions not captured in the continuum
Vanadium(4+)
DGs*ol=kJ molꢁ1
Vanadium(5+)
DGs*ol=kJ molꢁ1 model used, or atypical thermodynamic properties of its crystal
lattice. While no theoretical framework could account for all
possible physical properties of a real system, the high solubility of
CaVBH in DMSO underscores the importance of a predictive
theoretical framework, not to facilitate rational electrolyte design
[N1111]2[VBH]
[N2222]2[VBH]
[N3333]2[VBH]
[N4444]2[VBH]
ꢁ1111
ꢁ1065
ꢁ1049
ꢁ1054
[N1111][VBH]
[N2222][VBH]
[N3333][VBH]
[N4444][VBH]
ꢁ477
ꢁ454
ꢁ446
ꢁ449
but also to identify potentially valuable outliers.
Solubility trend for the oxidized, [VBH]1ꢁ species. A cycling
ow battery must comprise stable electrolytes in both oxidation
Fig. 6 Electrostatic potential maps plotted using natural charges of
1111, N2222, N3333 and N4444 cations (a–d respectively). Gray-shaded
region represents the polarizable continuum used by the SMD
method.
N
Fig. 7 Thermodynamic cycle for dissolution of VBH compounds of
N1111, N2222, N3333 and N4444 cations.
5436 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
states of the active-materials. As such, additional insights were as a result of lattice destabilization and despite a less favorable
gained by measuring the solubility of oxidized, [N3333][VBH] and solvation energy. Further, DG*sol and DGl*at change upon oxida-
[N4444][VBH] (Table 1) and these were incorporated into the tion, and in some cases are close in magnitude. Detailed
theoretical framework discussed above.
computational investigations of the lattice enthalpy of these
First, compared to reduced, [VBH]2ꢁ, one would expect compounds is currently underway to identify structural aspects
oxidized, [VBH]1ꢁ would exhibit a less stable crystal lattice (i.e., of the cation and VBH that may be changed to destabilize DG*lat
less negative DG*lat), due to the decreased overall charge of the and increase solubility. Finally, by considering the variability of
complex.48 A destabilized lattice should have the effect of DGs*ol for oxidized and reduced VBH-complexes with various
increasing solubility upon oxidation. Second, with respect to cations and in several solvents (see Fig. 8) we demonstrate that
the free energy of the solvated ions ðDG*solÞ, the loss of one a pathway exists to optimize both thermodynamic quantities for
cation, and the decreased charge on the anion should result in solubility by altering cation and anion structure, as well as by
a less stable solution (less negative DG*sol) for the oxidized, altering the dielectric constant of the solvent/solvent mixture
VBH1ꢁ complex compared to the reduced, [VBH]2ꢁ species. This used to formulate the electrolyte.
agrees with the DFT-calculated values of DG*sol for the oxidized
complexes (Fig. 8), which are less negative by 620–650 kJ molꢁ1
Electrochemical characterization of high-concentration VBH
in the oxidized state. A destabilized solution should have the
effect of decreasing solubility. Taken together with the experi-
mentally observed solubility, (see Table 1), several conclusions
can be drawn:
electrolytes
A critical aspect of improving solubility of VBH electrolytes is
maintaining high stability and cyclability by avoiding delete-
rious side-reactions that may become more severe at high
concentrations. As shown in Fig. 9a, [N4444]2[VBH] can be cycled
exhaustively at 0.24 mol Lꢁ1 to high depth-of-charge (DOC) with
no chemical decomposition, as monitored by UV-Vis spectros-
copy. Aer 16 round-trip cycles, comprising two initial and two
nal cycles between 90% and 10% DOC, with intervening cycles
between 60% and 40% DOC, virtually no chemical decomposi-
tion of the active-material was observed. Subsequently, elec-
trolyte solutions were remixed and cycled for an additional 6
cycles without observable decomposition. Notably, these
experiments took place over the course of more than one week
of constant cycling. As shown in Fig. 9b–g, the vanadium(4+)
and vanadium(5+) concentrations at each electrode can be
extracted by simulating the observed spectrum as a linear
combination of the reference spectra (see Fig. S2†). The spec-
troscopic signatures of the electrolytes are identical before and
aer the experiment and the total VBH concentration was
unchanged aer 183 h of galvanostatic cycling at between 5 and
8 mA. The long duration of each cycle, due to high area-specic
resistance, prevented investigation of active material-
concentrations near the limit of their solubility. We expect
that higher concentration studies will be more feasible at higher
current density, in a ow cell architecture.
ꢃUpon oxidation of [VBH]2ꢁ, the positive change in DG*sol upon
oxidation is greater than the positive change in DG*lat, otherwise
the active-material would become more soluble, not less.
ꢃSince the change in DG*sol upon oxidation of [N3333]2[VBH]
and [N4444]2[VBH] is nearly identical (ꢀ620 kJ molꢁ1), and
considering [N3333][VBH] is much less soluble than [N4444
]
[VBH], the positive change in lattice energy upon oxidation
must be less for [N3333]2[VBH] than for [N4444]2[VBH].
ꢃSince the solubility of oxidized, [N4444][VBH], is similar to
that of reduced, [N4444]2[VBH], the change in lattice energy
similar to the change in solvation energy, on the order of
600 kJ molꢁ1
.
In summary, the increase in solubility observed for VBH
compounds with longer-chain alkylammonium cations comes
Kinetics of high-concentration reduced/oxidized VBH by
ultramicroelectrode voltammetry
In addition to exhaustively cycling high-solubility VBH electro-
lytes, the kinetic behavior of these electrolytes at high concen-
trations was investigated. Cyclic voltammetry (CV) is a useful
tool to explore the fundamental electrochemical properties of
the electrolyte at low active-material concentrations using
Randle–Sevcik and Butler–Volmer theoretical models. However,
at higher concentrations, uncompensated resistances would
result in a deviation from the linear relationship, leading to
inaccurate predictions of the intrinsic electrochemical proper-
ties.49,50 To eliminate this problem, a carbon ber ultra-
microelectrode (UME) was used to probe the redox reactions at
near-iR-free conditions.50 Fig. 10 shows steady-state
Fig. 8 DFT-calculated values of DG*sol for reduced and oxidized
[NXXXX]y[VBH] compounds in various solvents with dielectric constant
varying between 2 and 80 (X ¼ 1 to 4, and Y ¼ 2 for reduced species
and Y ¼ 1 for oxidized species).
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5437

View Article Online
RSC Advances
Paper
Fig. 9 Electrochemical cycling of 0.24 mol Lꢁ1 [VBH] solution in MeCN, over the course of more than 180 hours, comprising 22 roundtrip cycles
of 80% and 20% DOC (a). Panels (b)–(g) correspond to UV-Vis traces and simulated traces used to determine concentration of VBH in the
reduced and oxidized states at the time point indicated by symbols in panel (a). Data were collected with a 20 mL 2-electrode cell with double-
sided MPL carbon cloth (12.9 cm2) current collector, with between 5 and 9 mA of current.
voltammograms of both reduced and oxidized forms of density increases and the diffusion coefficient decreases with
[N4444]2VBH in MeCN and the corresponding kinetic parameters increasing [N4444]2VBH concentration, while the observed rate
are summarized in Table 3. As expected, the exchange current constant decreases slightly. Although the oxidized form of VBH
Fig. 10 Steady state voltammograms of (a) reduced form, (b) oxidized form of [N4444]2VBH in MeCN at a scan rate of 0.05 V sꢁ1. All experiments
were performed using a carbon fiber UME at room temperature (25 ꢄC).
5438 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
Table 3 Electrochemical parameters of VBH electrolytes in MeCN, obtained using UME voltammetry, with j0 as exchange current density, k0 as
standard rate constant and D0 as the diffusion coefficient
Reduced form – [VBH]2ꢁ
Oxidized form – [VBH]1ꢁ
Conc. [M]
j0 [mA cmꢁ2
]
k0 ꢂ 10ꢁ3 [cm sꢁ1
]
D0 ꢂ 10ꢁ6 [cm2 sꢁ1
]
j0 [mA cmꢁ2
]
k0 ꢂ 10ꢁ3 [cm sꢁ1
]
D0 ꢂ 10ꢁ6 [cm2 sꢁ1
]
0.010
0.025
0.050
0.100
10.6
27.5
46.9
108
10.9
5.69
4.86
5.61
15.1
7.97
7.00
6.08
5.01
14.3
29.6
48.8
5.2
8.29
4.04
3.78
3.49
2.96
3.07
2.53
electrolytes exhibit lower electrochemical rate constants in of 1 cm path length. Calibration curves (see Fig. S4 and S5.†)
comparison to the reduced form due to increased viscosity (see were drawn to estimate the molar absorptivity of VBH1ꢁ
Fig. S3.†), this UME analysis demonstrates that high-solubility (reduced) and VBH2ꢁ (oxidized) at 825 nm and 485 nm
VBH electrolytes exhibit fast electron transfer kinetics with respectively. The total concentration and fraction of each redox
rate constants in the range of 10ꢁ2–10ꢁ3 cm sꢁ1
.
state of active material during cell cycling were estimated by
simulating the observed spectra as a linear combination of
reference spectra. Infrared spectra were recorded on a Thermo
Scientic Nicolet iS5 equipped with iD7ATR module and a dia-
mond crystal. 1H- and 13C-NMRs were recorded on Bruker
AVANCE III HD 400 MHz High-Performance Digital NMR
spectrometer operating at 400 MHz for 1H NMR, and 101 MHz
for 13C NMR. Data acquisition was performed on IconNMR
5.0.3, and spectra were processed in TopSpin 3.5 and Mnova.
High Resolution Mass Spectrometry (HRMS) were recorded on
a Waters Xevo QTOF-I spectrometer using direct infusion-
electrospray ionization technique. X-ray crystallographic
experiments were performed on a Bruker D8 Venture X-
instrument, using Mo Ka radiation at 200 K. Data were cor-
rected for absorption using SADABS. The structures were solved
by direct methods. All non-hydrogen atoms were rened
anisotropically by full matrix least squares on F2 and all
hydrogen atoms except those on water were placed in calculated
positions with appropriate riding parameters. Further rene-
ment and molecular graphics were obtained using Bruker Suite
of structural programs,53 and OLEX2.54
51,52
Conclusion
The increased solubility reported herein is an inroad to over-
coming a major obstacle to formulation of high energy-density
NRFB electrolytes with VBH-based active-materials. These
systems have the potential to address a growing need for ex-
ible, robust grid energy storage systems. In addition, the
insights developed herein provide a framework for electrolyte
optimization. Future efforts will focus on tailoring electrolyte
thermodynamics with combined theoretical and experimental
approaches and assessing ow cell performance using electro-
lytes with high active-material concentrations.
Experimental
General
Tetraethylammonium uoride hydrate (TEAF, 97%), tetra-n-
propylammonium chloride (TPACl, 99%), chloroacetic acid
(HClOAC, 98%), vanadium(IV) oxide acetylacetonate, and anhy-
drous potassium uoride (KF, 99%) were purchased from Bean
Town Chemicals. Tetra-n-butylammonium uoride hydrate
(TBAF) and tetra-n-butylammonium hexauorophosphate
(TBAPF6) were purchased from Oakwood Chemicals. TBAPF6
was recrystallized from ethanol twice prior to use. Hydroxyl-
amine hydrochloride (NH2OH$HCl, 99%) was purchased from
Alfa Aesar. Zinc acetate dihydrate, 97% was purchased from
Acros Organics. Tetramethylammonium uoride tetrahydrate
(TMAF, 98%) was purchased from Matrix Scientic. Calcium
chloride dihydrate was purchased from VWR. Ferrocenium
hexauorophosphate, 97% was purchased from Sigma Aldrich.
All chemicals were used as obtained unless mentioned other-
wise. All solvents were obtained from Fisher Scientic, distilled
over drying reagent and stored in a nitrogen-lled glovebox with
oxygen and moisture content less than 1 ppm.
Electrochemical methods
All electrochemical experiments were carried out in N2-lled
glovebox with oxygen and water content less than 1 ppm.
Cyclic voltammetry was performed in a custom designed cell.
Working electrode was 2 mm diameter glassy carbon electrode.
Platinum wire was used as counter electrode. Nonaqueous
reference electrode was fabricated by submerging a piece of
silver wire in 25 mM silver nitrate solution in acetonitrile with
0.1 M TBAPF6 in a glass tube tted with vycor frit on one end.
CV and charge discharge experiments were recorded using
Ametek VersaSTAT3 (Princeton Applied Research).
Ultramicroelectrode (UME) voltammetry: analysis of the
redox reaction kinetics in high-concentration VBH electrolytes
was performed with a UME voltammetry setup. 11 mm-diameter
carbon ber UME (BASi) was used as the working electrode.
Platinum wire and silver wire were selected as counter and
Physical methods
UV-Vis spectra were collected in a Evolution 220 UV-visible reference electrodes, respectively. The working electrode was
Spectrophotometer (Thermo Scientic) using a quartz cuvette prepared by rinsing the 11 mm carbon ber with DI H2O,
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5439

View Article Online
RSC Advances
Paper
polishing with 0.05 mm alumina suspension, and then rinsing stirrer. Hydrochloric acid (80.0 mL, 6.0 M) was added dropwise
a second time with DI H2O. The reference electrode was cali- to the mixture with stirring until dissolution of all the reactants
brated against ferrocene in the respective support electrolyte. occurred. Calcium chloride dihydrate (28.8 g, 0.197 mol) was
5 mL of electrolyte were used for all UME experiments.
added to the mixture and dissolved. CaVBH was precipitated
The standard rate constants (k0) were calculated according to from resulting dark blue solution by addition of 2 L of iso-
the Butler–Volmer equation:
propanol. The product was ltered and washed with 20 mL
isopropanol followed by acetone and dried in vacuo for 24 hours
at room temperature. (Yield: 78.87 g, 0.167 mol, 85%) UV-Vis
(DMSO): lmax ¼ 578 nm (3 ¼ 27.5 molꢁ1 cmꢁ1). IR (n/cmꢁ1):
3541w, 3321w, 2988w, 1589 s (C]O).
I0
k0 ¼
nFAC
where I0 is exchange current (A), n is the number of electrons, F
is the Faraday constant (96485 C molꢁ1), A is the electrode area,
and C is the active species concentration (mol cmꢁ3). Quanti-
tative analysis of diffusion coefficients (D0) of was derived from
the steady-state current plateau of the UME voltammograms.
The steady-state current is used to calculate D0 using the
following relationship:
[(N1111)2VBH]. CaVBH (4.32 g, 0.00913 mol) was added to
stirring mixture of tetramethylammonium uoride tetrahydrate
(3.01 g, 0.018 mol) in 50 mL ethanol. The reaction mixture was
stirred at 30 ꢄC for 1 h. The solution of [(N1111)2VBH] in ethanol
was isolated by centrifuging mixture at 3000 rpm for 10 min and
concentrated using a rotatory evaporator. Crystals were ob-
tained by cooling the solution in an ice bath, collected via
vacuum ltration, and dried in vacuo for 16 h at room temper-
ature. (Yield: 3.40 g, 0.00607 mol, 66%) Single crystals suitable
for X-ray analysis were obtained by diffusing THF into solution
of compound in ethanol. UV-Vis (DMSO): lmax ¼ 579 nm (3 ¼
30.2 molꢁ1 cmꢁ1). IR (n/cmꢁ1): 3345w, 3031w, 2962w (C–H),
1619s (C]O), 1492w (C–N). HR-MS (ESI, positive ion mode) m/z
calculated for C4H12N+ [M]+: 74.0964, found: 74.0954.
[(N2222)2VBH]. CaVBH (9.60 g, 0.0203 mol) was added to
solution of tetraethylammonium uoride (6.11 g, 0.0409 mol) in
60.0 mL acetonitrile and stirred overnight at room temperature.
CaF2 precipitate was separated by vacuum ltration and the
ltrate volume was reduced to ca. 20 mL in a rotatory evaporator
which on cooling yielded blue crystals. Product was separated,
washed with cold THF and dried in vacuo. (Yield, 11.3 g,
0.0182 mol, 90%). UV-Vis (MeCN): lmax ¼ 578 nm (3 ¼ 28.2
molꢁ1 cmꢁ1). IR (n, cmꢁ1): 3506w, 2994w, 2955w, 2909w, 1602s
(C]O). HR-MS (ESI, positive ion mode) m/z calcd for C8H20N+
[M]+: 130.1590, found: 130.1554.
iss ¼ 4FrD0nC
where F is Faraday constant and r is the radius of the electrode
(5.5 mm). C indicates the concentration of redox materials.
Galvanostatic charge discharge cycling was conducted with
a two-electrode cell setup in a 20 mL H-cell from Adams and
Chittenden Scientic. Double sided MPL carbon cloth (12.9
cm2) was used as electrode. Naon membrane was soaked in
solution of 0.5 M TBAPF6 in MeCN for 24 h prior to use.
Solubility of alkylammonium [VBH] compounds
Solubilities were experimentally measured by UV-Vis absor-
bance spectroscopy at room temperature, based on calibration
curves (see Fig. S4 and S5.†). Saturated solutions of each of the
compounds were made in the four different solvents and
centrifuged. Samples were then diluted for measurement.
Synthetic methods
Ligand synthesis. Zinc(2+) hydroxyiminodiacetate (ZnHIDA)
[(N3333)2VBH]. [(N3333)2VBH] was synthesized in the same
was synthesized using literature method55 with modication as manner as [(N1111)2VBH], using CaVBH and tetrapropylammo-
follows. Hydroxylamine hydrochloride (70.1 g, 1.01 mol) was nium uoride (TPAF). Tetrapropylammonium uoride (TPAF)
neutralized with 5 M NaOH (200 mL, 1.00 mol) in an ice bath was prepared by halogen exchange, as reported previously.56
while maintaining the temperature of the reaction mixture Tetrapropylammonium chloride (12.7 g, 0.0574 mol) was dis-
below 4 ꢄC. In another ask, chloroacetic acid (189 g, 2.00 mol) solved in 30 mL dry methanol. KF (4.10 g, 0.0706 mol, 1.25 equ.)
was slowly neutralized with 5 M NaOH (400 mL, 2.00 mol) in an and 160 mL water (4% by weight of KF) was added to the solution
ice bath. Neutralized solution of chloroacetic acid was then and the mixture was stirred for 15 min. and ltered. The process
added dropwise to hydroxylamine solution. An additional was repeated twice and the nal ltrate was reduced to volume
400 mL 5 M NaOH solution was added to the reaction mixture of ca. 7 mL which was ltered to obtain a yellow, oily liquid. A
and was stirred for 72 h in an ice bath. The reaction mixture was suspension of CaVBH (7.80 g, 0.0165 mol) in 40 mL acetonitrile
then and acidied to pH 3.9 by adding 6 M HCl. Zinc acetate was prepared to which above yellow oil was added and stirred,
dihydrate (220 g, 1.00 mol) was then added to the solution while resulting in an immediate dissolution of suspended CaVBH.
stirring. White precipitate of ZnHIDA was ltered and washed Additional CaVBH (1.50 g, 0.00317 mol) was added to the
with cold water several times and dried in vacuo. Yield: 201 g, solution in increments of 0.5 g aer which further addition
0.948 mol, 95%. IR (n/cmꢁ1): 3228w, 2930w (C–H), 1648 m, would result in insoluble CaVBH indicating the molar equiva-
1576s (C]O).
lence point. The reaction mixture was centrifuged at 4000 rpm
Synthesis of active materials. CaVBH was synthesized using for 15 min. to remove precipitate from the blue liquid. Crystals
a modication of literature procedures55 with slight modica- were obtained by treating the cold solution of the compound
tion for scale up. ZnHIDA (83.2 g, 0.392 mol) was added to with THF which was re-crystalized from MeCN/THF but a suffi-
500 mL de-ionized water to which vanadyl(4+)acetylacetonate ciently high-resolution X-ray crystal structure was not obtained
(52.0 g, 0.196 mol) was added and mixed well using magnetic due to crystallographic disorder or other crystallographic issue.
5440 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
(Yield: 14.0 g, 0.0196 mol, 60%). UV-Vis (MeCN): lmax ¼ 579 nm v4.2.1.57 The Perdew–Burke–Ernzerhof (PBE) generalized
(3 ¼ 27.9 molꢁ1 cmꢁ1). IR (n, cmꢁ1): 3471w, 2970w (C–H), 2878w, gradient approximation (GGA) functional and Grimme's DFT-
1607s (C]O), 1488w (C–N). HR-MS (ESI, positive ion mode) m/z D3 dispersion correction with Becke–Johnson damping (D3BJ)
calcd for C12H28N+ [M]+: 186.2216, found: 186.2242.
using the def2-TZVP basis set was used in the equilibrium
[(N4444)2VBH]. [(N4444)2VBH] was prepared in the same geometry optimizations in both gaseous and solvated
manner as [(N1111)2VBH], using tetrabutylammonium-uoride systems.58–62 The DGs*ol for each ion was then taken as the
trihydrate (12.9 g, 0.0411 mol) and CaVBH (9.71 g, 0.0205 difference between the solvated total energy (Esol) and the
mol) in 40 mL THF (yield: 14.1 g, 0.0170 mol, 83%). Single gaseous total energy (Evac) (eqn (2)) and corrected for the free
crystals were obtained by slow diffusion of diethyl ether into energy change (DG0/* ¼ 1.9 kcal molꢁ1 at 298 K) associated in
solution of the product in MeCN. UV-Vis (MeCN): lmax ¼ 577 nm converting 1 atm of gas to 1 mol Lꢁ1 solution (eqn (3)). Eqn (2)
(3 ¼ 28.1 molꢁ1 cmꢁ1). IR (n, cmꢁ1): 3419w, 2957w (C–H), 2872w, holds within the SMD method as this continuum model is
1615s (C]O), 1490w (C–N). HR-MS (ESI, positive ion mode) m/z parameterized with a substantial dataset of experimental free
calcd. for C16H36N+ [M]+: 242.2842, found: 242.2829.
energies of solvation at room temperature.
[(N4444)VBH]. [(N4444)2VBH] was chemically oxidized to
[(N4444)VBH] using ferrocenium hexauorophosphate (FcPF6).
In a Schlenk ask [(N4444)2VBH] (7.28 g, 0.00880 mol) was dis-
solved in 30 mL acetonitrile. Suspension of FcPF6 (2.91 g,
0.00879 mol) in acetonitrile was added to the ask and stirred
for 15 min. The solution was ltered, and the ltrate was roto-
vapped to reduce the volume to ca. 15 mL. The product was
crystallized from acetonitrile/diethyl ether and repeatedly
washed with hexane and ether until ferrocene was removed. Red
crystalline product was isolated and dried under high vacuum
for 24 h. Single crystal suitable for X-ray analysis was obtained
by slow diffusion of diethyl ether into solution of the title
compound in MeCN (yield: 5.00 g, 0.00855 mol, 97%). UV-Vis
(MeCN): lmax ¼ 497 nm (3 ¼ 240 molꢁ1 cmꢁ1). IR (n, cmꢁ1):
2962w (C–H), 2875w, 1667s (C]O), 1484w (C–N). dH (400 MHz,
CD3CN) 4.94 (2H, d, J 16.1), 4.63 (4H, dd, J 38.4, 16.1), 4.41 (2H,
d, J 16.1), 3.08 (8H, t, J 7.3), 1.60 (8H, t, J 7.4), 1.35 (8H, q, J 7.2),
0.96 (12H, t, J 7.3). dC (101 MHz, CD3CN) 171.00 (d, J 41.3), 65.56
(d, J 67.8), 59.40, 24.38, 20.41, 13.87.
DG0sol ¼ Esol ꢁ Evac
(2)
(3)
DGs*ol ¼ DGs0ol þ DG0/*
Electrostatic potential plots were prepare using the Multiwfn
soware package,63 incorporating natural charges calculated
from a natural bond orbital analysis,43 and cube les were
visualized using the Chemcra soware package.64
Conflicts of interest
There are no conicts to declare.
Acknowledgements
This research was supported in part by UMass Dartmouth's
[(N3333)VBH]. [(N3333)VBH] was prepared by chemical oxida- Marine and Undersea Technology (MUST) Research Program
tion of [(N3333)2VBH] solution in MeCN (14.0 mL, 0.576 M, funded by the Office of Naval Research (ONR) under Grant No.
0.00806 mol) using FcPF6 (2.66 g, 0.00804 mol). FcPF6 was N00014-20-1-2170. This work was also supported in part by
directly added to the solution and stirred for 30 min. Reddish UMass through the OTCV Technology Development Fund. P. J.
brown precipitate was isolated and washed with toluene C. and M. L. M. gratefully acknowledge UMass Dartmouth and
multiple times until trace of ferrocene was removed. Product the UMass Green High-Performance Computing Cluster for
was recrystallized from a MeCN/diethyl ether mixture. Red computing resources. This material is also based upon work
powder was isolated via ltration and dried in Schlenk line. supported by the U.S. Department of Energy, Office of Science,
(Yield 3.90 g, 0.00737 mol, 91%) crystals suitable for X-ray under Award Number DE-SC0020585 (Prime contractor, Triton
diffraction were grown by vapour diffusion of diethyl ether Systems Inc). This report was prepared as an account of work
into acetonitrile solution of the title compound. UV-Vis (MeCN): sponsored by an agency of the United States Government.
lmax ¼ 495 nm (3 ¼ 239.88 molꢁ1 cmꢁ1). IR (n, cmꢁ1): 2968w (C– Neither the United States Government nor any agency thereof,
H), 2880w, 1666s (C]O), 1486w (C–N). dH (400 MHz, CD3CN) nor any of their employees, makes any warranty, express or
4.94 (2H, d, J 16.1), 4.63 (4H, dd, J 38.2, 16.1), 4.41 (2H, d, J 16.1), implied, or assumes any legal liability or responsibility for the
3.11–2.97 (8H, m), 1.73–1.56 (8H, m), 0.94 (12H, t, J 7.3).
accuracy, completeness, or usefulness of any information,
apparatus, product, or process disclosed, or represents that its
use would not infringe privately owned rights. Reference herein
to any specic commercial product, process, or service by trade
Computational methods
The solvation free energies ðDG*solÞ of the +4 and +5 oxidation name, trademark, manufacturer, or otherwise does not neces-
states of the VBH anion with the corresponding Nxxxx (x ¼ sarily constitute or imply its endorsement, recommendation, or
1,2,3,4 is the number of alkyl carbon atoms) alkylammonium favoring by the United States Government or any agency thereof.
cations were investigated over a range of solvent systems (Table The views and opinions of authors expressed herein do not
S1†).
necessarily state or reect those of the United States Govern-
Implicit solvation was induced using the solvation model ment or any agency thereof. The authors also thank Louis
based on density (SMD) method implementation in Orca Carreiro for his helpful insights on this work.
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5441

View Article Online
RSC Advances
Paper
21 Y. Yan, T. P. Vaid and M. S. Sanford, J. Am. Chem. Soc., 2020,
142, 17564–17571.
References
22 L. E. VanGelder, B. E. Petel, O. Nachtigall, G. Martinez,
W. W. Brennessel and E. M. Matson, ChemSusChem, 2018,
11, 4139–4149.
23 J. M. Stauber, S. Zhang, N. Gvozdik, Y. Jiang, L. Avena,
K. J. Stevenson and C. C. Cummins, J. Am. Chem. Soc.,
2018, 140, 538–541.
24 J. Friedl, M. V. Holland-Cunz, F. Cording, F. L. Pfanschilling,
C. Wills, W. McFarlane, B. Schricker, R. Fleck,
H. Wolfschmidt and U. Stimming, Energy Environ. Sci.,
2018, 11, 3010–3018.
1 Sustainable Energy in America Factbook, https://www.bcse.org/
download/, accessed April 11, 19.
2 P. Albertus, J. S. Manser and S. Litzelman, Joule, 2020, 4, 21–
32.
3 U. S. E. I. Administration, Levelized Cost and Levelized Avoided
Cost of New Generation Resources AEO20201 February 2020
Levelized Cost and Levelized Avoided Cost of New Generation
Resources in the Annual Energy Outlook 2020, https://
www.eia.gov/outlooks/aeo/pdf/electricity_generation.pdf,
(accessed 7/16/20, 2020).
25 P. B. Tsitovich, A. M. Kosswattaarachchi, M. R. Crawley,
T. Y. Tittiris, T. R. Cook and J. R. Morrow, Chemistry, 2017,
23, 15327–15331.
26 Y. Li, J. Sniekers, J. Malaquias, X. Li, S. Schaltin, L. Stappers,
K. Binnemans, J. Fransaer and I. F. J. Vankelecom,
Electrochim. Acta, 2017, 236, 116–121.
27 Y. Ding, Y. Zhao, Y. T. Li, J. B. Goodenough and G. H. Yu,
Energy Environ. Sci., 2017, 10, 491–497.
28 P. J. Cappillino, H. D. Pratt, N. S. Hudak, N. C. Tomson,
T. M. Anderson and M. R. Anstey, Adv. Energy Mater., 2014,
4, 1300566.
4 ERCOT - Grid Information - Generation, http://www.ercot.com/
gridinfo/generation, (accessed 7/16/20, 2020).
5 M. Ulaganathan, V. Aravindan, Q. Yan, S. Madhavi,
M. Skyllas-Kazacos and T. M. Lim, Adv. Mater. Interfaces,
2016, 3, 1500309.
6 B. Dunn, H. Kamath and J.-M. Tarascon, Science, 2011, 334,
928–935.
7 C. Zhang, L. Zhang, Y. Ding, S. Peng, X. Guo, Y. Zhao, G. He
and G. Yu, Energy Storage Mater., 2018, 15, 324–350.
8 K. Gong, Q. Fang, S. Gu, S. F. Y. Li and Y. Yan, Energy Environ.
Sci., 2015, 8, 3515–3530.
29 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, J. Mater. Chem. A, 2015, 3, 7929–7938.
30 Y. Y. Lai, X. Li and Y. Zhu, ACS Appl. Polym. Mater., 2020, 2,
113–128.
9 R. W. Hogue and K. E. Toghill, Curr. Opin. Electrochem., 2019,
18, 37–45.
10 Y. Yan, S. G. Robinson, M. S. Sigman and M. S. Sanford, J.
Am. Chem. Soc., 2019, 141, 15301–15306.
11 N. H. Attanayake, J. A. Kowalski, K. V. Greco,
M. D. Casselman, J. D. Milshtein, S. J. Chapman,
S. R. Parkin, F. R. Brushett and S. A. Odom, Chem. Mater.,
2019, 31, 4353–4363.
12 K. H. Hendriks, S. G. Robinson, M. N. Braten, C. S. Sevov,
B. A. Helms, M. S. Sigman, S. D. Minteer and
M. S. Sanford, ACS Cent. Sci., 2018, 4, 189–196.
13 J. Friedl, M. A. Lebedeva, K. Porfyrakis, U. Stimming and
T. W. Chamberlain, J. Am. Chem. Soc., 2018, 140, 401–405.
14 J. Winsberg, T. Hagemann, T. Janoschka, M. D. Hager and
U. S. Schubert, Angew. Chem., Int. Ed., 2017, 56, 686–711.
15 P. Leung, A. A. Shah, L. Sanz, C. Flox, J. R. Morante, Q. Xu,
´
31 E. C. Montoto, G. Nagarjuna, J. S. Moore and J. Rodrıguez-
´
Lopez, J. Electrochem. Soc., 2017, 164, A1688–A1694.
32 R. M. Darling, K. G. Gallagher, J. A. Kowalski, S. Ha and
F. R. Brushett, Energy Environ. Sci., 2014, 7, 3459–3477.
33 H. Huang, R. Howland, E. Agar, M. Nourani, J. A. Golen and
P. J. Cappillino, J. Mater. Chem. A, 2017, 5, 11586–11591.
34 T. C. Gokoglan, S. K. Pahari, A. Hamel, R. Howland,
P. J. Cappillino and E. Agar, J. Electrochem. Soc., 2019, 166,
A1745–A1751.
35 J. F. Kucharyson, L. Cheng, S. O. Tung, L. A. Curtiss and
L. T. Thompson, J. Mater. Chem. A, 2017, 5, 13700–13709.
36 L. Cheng, P. Redfern, K. C. Lau, R. S. Assary, B. Narayanan
and L. A. Curtiss, J. Electrochem. Soc., 2017, 164, E3696–
E3701.
´
M. R. Mohamed, C. Ponce de Leon and F. C. Walsh, J.
Power Sources, 2017, 360, 243–283.
16 J. A. Kowalski, M. D. Casselman, A. P. Kaur, J. D. Milshtein,
C. F. Elliott, S. Modekrutti, N. H. Attanayake, N. Zhang,
S. R. Parkin, C. Risko, F. R. Brushett and S. A. Odom, J.
Mater. Chem. A, 2017, 5, 24371–24379.
37 R. E. Skyner, J. L. McDonagh, C. R. Groom, T. van Mourik
and J. B. O. Mitchell, Phys. Chem. Chem. Phys., 2015, 17,
6174–6191.
38 S.-J. Lu, H.-H. Yang, W.-J. Chang, H.-H. Hsueh, Y.-C. Lin,
F.-C. Liu, I. J. B. Lin and G.-H. Lee, J. Organomet. Chem.,
2020, 927, 121543.
39 F. E. Hahn, M. C. Jahnke and T. Pape, Organometallics, 2007,
26, 150–154.
40 E. M. Armstrong, D. Collison, N. Ertok and C. D. Garner,
Talanta, 2000, 53, 75–87.
41 A. V. Marenich, C. J. Cramer and D. G. Truhlar, J. Phys. Chem.
B, 2009, 113, 6378–6396.
17 J. D. Milshtein, J. L. Barton, R. M. Darling and F. R. Brushett,
J. Power Sources, 2016, 327, 151–159.
18 T. B. Liu, X. L. Wei, Z. M. Nie, V. Sprenkle and W. Wang, Adv.
Energy Mater., 2016, 6, 1501449.
19 G.-M. Weng, B. Yang, C.-Y. Liu, G.-Y. Du, E. Y. Li and
Y.-C. Lu, Energy Environ. Sci., 2019, 12, 2244–2252.
20 J. D. Milshtein, A. P. Kaur, M. D. Casselman, J. A. Kowalski,
S. Modekrutti, P. L. Zhang, N. Harsha Attanayake,
C. F. Elliott, S. R. Parkin, C. Risko, F. R. Brushett and
S. A. Odom, Energy Environ. Sci., 2016, 9, 3531–3543.
42 J. R. Rumble, CRC Handbook of Chemistry and Physics, CRC
Press, 101st edn, 2020.
5442 | RSC Adv., 2021, 11, 5432–5443
© 2021 The Author(s). Published by the Royal Society of Chemistry

View Article Online
Paper
RSC Advances
43 E. D. Glendening, C. R. Landis and F. Weinhold, J. Comput. 53 Bruker, APEX3, SAINT, and SADABS, Bruker AXS Inc.,
Chem., 2013, 34, 1429–1437. Madison, Wisconsin, USA, 2008.
44 J. Łubkowski and J. Bła˙zejowski, Thermochim. Acta, 1990, 54 O. V. Dolomanov, L. J. Bourhis, R. J. Gildea, J. A. K. Howard
157, 259–277. and H. Puschmann, J. Appl. Crystallogr., 2009, 42, 339–341.
45 P. Dokurno, J. Łubkowski and J. Błaejowski, Thermochim. 55 G. Anderegg, E. Koch and E. Bayer, Inorg. Chim. Acta, 1987,
Acta, 1990, 165, 31–48. 127, 183–188.
46 J. Błazejowski and E. Kowalewska, Thermochim. Acta, 1986, 56 S. Dermeik and Y. Sasson, J. Org. Chem., 1989, 54, 4827–4829.
˙
˙
105, 257–286.
57 F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2,
73–78.
58 A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
47 I. Krossing and I. Raabe, Angew. Chem., Int. Ed., 2004, 43,
2066–2090.
48 P. Atkins and J. De Paula, Physical Chemistry, W.H. Freeman 59 A. D. Becke, J. Chem. Phys., 1992, 97, 9173–9177.
and Company, New York, 7 edn, 2002.
49 J. A. Kowalski, A. M. Fenton Jr, B. J. Neyhouse and
F. R. Brushett, J. Electrochem. Soc., 2020, 167, 160513.
60 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B, 1988, 37, 785–
789.
61 A. D. Becke, Phys. Rev. A, 1988, 38, 3098–3100.
50 L. Su, M. Ferrandon, J. A. Kowalski, J. T. Vaughey and 62 S. H. Vosko, L. Wilk and M. Nusair, Can. J. Phys., 1980, 58,
F. R. Brushett, J. Electrochem. Soc., 2014, 161, A1905–A1914. 1200–1211.
51 A. M. Kosswattaarachchi, L. E. VanGelder, O. Nachtigall, 63 T. Lu and F. Chen, J. Comput. Chem., 2012, 33, 580–592.
J. P. Hazelnis, W. W. Brennessel, E. M. Matson and 64 G. A. Andrienko, Chemcra graphical soware for
-
T. R. Cook, J. Electrochem. Soc., 2019, 166, A464–A472.
52 R. W. Hogue, C. G. Armstrong and K. E. Toghill,
ChemSusChem, 2019, 12, 4506–4515.
visualization of quantum chemistry computations, https://
www.chemcraprog.com.
© 2021 The Author(s). Published by the Royal Society of Chemistry
RSC Adv., 2021, 11, 5432–5443 | 5443