Cobalt and Vanadium Trimetaphosphate Polyanions:
Synthesis, Characterization, and Electrochemical Evaluation
for Non-aqueous Redox-Flow Battery Applications
The MIT Faculty has made this article openly available.
Please share
how this access benefits you. Your story matters.
Citation
Stauber, Julia M. et al. “Cobalt and Vanadium Trimetaphosphate
Polyanions: Synthesis, Characterization, and Electrochemical
Evaluation for Non-Aqueous Redox-Flow Battery Applications.”
Journal of the American Chemical Society 140, 2 (January 2018):
538–541 © 2017 American Chemical Society
As Published
http://dx.doi.org/10.1021/jacs.7b08751
Publisher
American Chemical Society (ACS)
Version
Author's final manuscript
Citable link
http://hdl.handle.net/1721.1/119623
Terms of Use
Article is made available in accordance with the publisher's
policy and may be subject to US copyright law. Please refer to the
publisher's site for terms of use.
Cobalt and Vanadium Trimetaphosphate Polyanions: Synthesis, Characterization
and Electrochemical Evaluation for Non-Aqueous Redox-Flow Battery Applications
Julia M. Stauber,†Shiyu Zhang,†Nataliya Gvozdik,‡Yanfeng Jiang,†Laura Avena,†Keith J. Stevenson,∗,‡and Christopher C. Cummins∗,†
Department of Chemistry, Massachusetts Institute of Technology, 77 Massachusetts Ave, Cambridge, MA 02139, and Center for Electrochemical Energy Storage, Skolkovo Institute of Science and Technology, Moscow 143026, Russia
Received December 4, 2017; E-mail: k.stevenson@skoltech.ru; ccummins@mit.edu
ABSTRACT: An electrochemical cell consisting of cobalt ([CoII/III(P3O9)2]4−/3−) and vanadium ([VIII/II(P3O9)2]3−/4−)
bistrimetaphosphate complexes as catholyte and anolyte species, respectively, was constructed with a cell voltage of 2.4 V and Coulombic efficiencies exceeding 90% for up to 100 total cy-cles. The [Co(P3O9)2]4− (1) and [V(P3O9)2]3− (2) complexes
have favorable properties for flow-battery applications that in-clude reversible redox chemistry, high stability toward electro-chemical cycling, and high solubility in MeCN (1.09 ± 0.02 M, [PPN]4[1]·2MeCN; 0.77 ± 0.06 M, [PPN]3[2]·DME). The
[PPN]4[1]·2MeCN and [PPN]3[2]·DME salts were isolated as
crystalline solids in 82 and 68% yields, respectively, and charac-terized by 31P NMR, UV-vis, ESI-MS(–), and IR spectroscopy. The [PPN]4[1]·2MeCN salt was also structurally characterized,
crystallizing in the monoclinic P21/c space group. Treatment of
1 with [(p-BrC6H4)3N]+allowed for isolation of the one-electron
oxidized spin-crossover complex, [Co(P3O9)2]3−(3), which is the
active catholyte species generated during cell charging. The pres-ence of a spin-crossover complex is a feature of the present system that may enable voltage window variation as a function of temper-ature. The success of the 1-2 cell provides a promising entry point to a potential future class of transition-metal metaphosphate-based all-inorganic non-aqueous redox-flow battery electrolytes.
Redox-flow battery (RFB) technologies have gained widespread interest as promising solutions for renewable and efficient grid-scale energy storage.1RFBs offer many advantages over traditional
redox storage solutions such as solid-electrode batteries due to their relatively low cost, high efficiency, high scalability, modularity, and longer lifetimes.1,2 Aqueous vanadium-based RFBs are the cur-rent state-of-the-art, and have successfully demonstrated reliable electrochemical performance at the commercial level.2–4 Despite their immense success, aqueous RFBs suffer from several disad-vantages that include low energy density (∼50 Whl−1), limited op-erational temperatures (0–100◦C), and a narrow electrochemical window (1.2 V).1,5To overcome these challenges, recent research is shifting to non-aqueous RFBs (NARFBs).6 NARFBs expand the operational voltage window from 1.2 V (water) to much wider electrochemical windows exceeding 5 V depending on the organic solvent employed;7–13 NARFBs also offer the potential to work outside of the typical operating temperature of water. However, traditional NARFBs that employ transition-metal complex elec-trolytes supported by organic ligands (dithiolate,3,14
cyclopenta-dienyl,15 bipyridine,16,17 acetylacetonate9,11) suffer from numer-ous drawbacks such as low cyclability,18 high production cost,
†Massachusetts Institute of Technology ‡Skolkovo Institute of Science and Technology
high flammability, limited electrochemical stability,8,18ligand
dis-sociation,8,13,19and cross-contamination between the anolyte and catholyte compartments;10these drawbacks have thus far prevented wide-scale adoption of NARFBs.
Herein, we report a new approach to NARFB electrolyte de-sign through the use of redox-active cyclic phosphate metal com-plexes. This work presents trimetaphosphate ([P3O9]3−) anions
as all-inorganic supporting ligands for cobalt and vanadium ions (Figure 1, B; [PPN]4[Co(P3O9)2]·2MeCN, [PPN]4[1]·2MeCN;
[PPN]3[V(P3O9)2]·DME, [PPN]3[2]·DME), in which the
redox-active metal center’s coordination sphere is completed entirely by a pair of metaphosphate rings. The electrochemical inertness of the redox inactive cyclic phosphates ([PO3]nn−), coupled with their
low cost, low toxicity, and flame-retardant nature20 render them promising building blocks for flow-battery chemistries. When 1 and 2 were paired as catholyte and anolyte species, respectively, the resulting dual-active-species electrochemical cell displayed a total cell voltage of 2.4 V. The chelating nature of the trimetaphos-phate ring aids in disfavoring ligand dissociation,13and the lack of C–H bonds provides oxidative stability to the redox-active species that increases the cell’s lifetime when compared with the cycle life-times of other NARFB candidate systems.9,11The transition-metal complexes are also polyanionic in all charge states that are relevant to cycling of the RFB, a property that minimizes crossover of ac-tive species through membrane separators. Additionally, the solu-bilities of the metaphosphate complexes in polar organic media are enhanced when compared with solubilities of other transition metal complex electrolytes due to the highly lipophilic PPN+(PPN+= bis(triphenylphosphine)iminium) cations. The cation, however, is a variable component of the current system. The PPN+cation was chosen for the present study on the basis of our previous work,21–23 but other cations could be selected to optimize properties including solubility and cost.
2 [PPN]2[P3O9H] [PPN]y[Mn+(P3O9)2] –n acacH Mn+(acac) n + n–2 H+ + 1, M = Co, n = 2, y = 4 2, M = V, n = 3, y = 3 (1) We have previously shown that monohydrogen trimetaphos-phate ([P3O9H]2−) is capable of effecting ligand exchange
through protonolyis of acetylacetonate (acac) from [OTi(acac)2]2
to generate the terminal titanyl trimetaphosphate complex, [OTiP3O9(acac)]2−.21 Employing this strategy of protolytic
ligand replacement, salts of complexes [Co(P3O9)2]4− (1)
and [V(P3O9)2]3− (2) were prepared through treatment of
[PPN]2[P3O9H] (2 equiv) with Co(acac)2 or V(acac)3 (1 equiv),
respectively (Equation 1). Acetylacetone (acacH), the byproduct formed through protonation, is easily separable from the transition-metal metaphosphate complexes, making the preparation of the PPN+salts of 1 and 2 convenient and straightforward. Preparation of [PPN]3[V(P3O9)2]·DME, however, requires three equivalents
Potential (V vs. Fc/Fc+) 1.0 0.5 0 -0.5 -1.0 -1.5 -2.0 -2.5 ΔE = 2.4 V [VII(P 3O9)2]4– [VIII(P 3O9)2]3– [CoII(P 3O9)2]4– [CoIII(P 3O9)2]3– 2 μA 1
A
B
4–/3– O P O O O O O O O O P O P CoII/III P O O O O O O O O P P P O O O O O O O O O P P VIII/II P O O O O O O O O O P P 3–/4––
+
PF6– 2 e– charging porous carbon electrode mono-anion selective membrane e– dischargingFigure 1. A: Cyclic voltammograms of 2 mM MeCN solutions of [PPN]4[1]·2MeCN (blue) and [PPN]3[2]·DME (gray) scanned at 100 mV/s and referenced vs. Fc/Fc+. Working electrode: glassy carbon, pseudo-reference electrode: Ag/Ag+, counter electrode: Pt wire. Arrows indicate scan direction. B: Schematic representation of an H-cell used for charge-discharge studies with molecular structure drawings of [Co(P3O9)2]4−/3−(catholyte, left), and [V(P3O9)2]3−/4− (anolyte, right). Typical experiments were run at 5 mM [PPN]4[1]·2MeCN and [PPN]3[2]·DME concentrations in a 0.5 M MeCN solution of [TBA][PF6] with a Neosepta® anion-selective membrane separating each compartment, and carbon paper electrodes.
of acid for protonation of all three acac ligands of V(acac)3. Two
equivalents are provided by [PPN]2[P3O9H], and the third is
pro-vided by addition of one equivalent of p-CF3C6H4COOH. The
p-CF3C6H4COOH reagent was chosen as an appropriate acid for
protonation of acac from V(acac)3based on pKavalues,24and due
to the formation of the [PPN][p-CF3C6H4COO] byproduct, which
is easily separated away from [PPN]3[2]·DME by thorough
wash-ing of the isolated material with THF. Both [PPN]4[1]·2MeCN
and [PPN]3[2]·DME were crystallized from MeCN/DME solvent
mixtures, and isolated as crystalline solids in 82 and 68% yields, respectively.
The identity of 1 as the bistrimetaphosphate cobalt(II) com-plex was confirmed by an X-ray diffraction study (SectionS12). The solid-state structure of 1 displays the cobalt(II) ion within an all-oxygen octahedral coordination environment provided by the two [P3O9]3−rings with a Co–Oavgdistance of 2.108(9) Å (
Fig-ure S64). An X-ray diffraction study of [PPN]3[2]·DME was not
conducted as the solid-state structure of [NEt4]3[2] has been
re-ported previously.25
Complex 1 is characterized by a31P NMR resonance located at δ 616.7 ppm (∆ν1/2= 466 Hz, MeCN, 25◦C,Figure S1) that is
assigned to the phosphorus atoms of the [P3O9]3− ligand. This
signal is shifted significantly downfield from that of monohydro-gen trimetaphosphate (δ –22.5 ppm, MeCN)23 due to the influ-ence imparted by the paramagnetic Co(II) d7 metal center. Con-sistent with the presence of a high-spin Co(II) ion, complex 1 dis-plays an S = 3/2 electronic configuration at 25◦C in CD3CN, as
determined by the Evans method (µeff = 4.10 µB).26 The
elec-tronic spectrum of 1 measured in MeCN (Figure S6, Figure S7) is characterized by absorptions located at λmax 487 (sh), 540 (ε
= 5 M−1cm−1), and 1135 (ε = 1 M−1cm−1), which are assigned to the4A2g(F)←4T1g(F),4T1g(P)←4T1g(F), and4T2g(F)←4T1g(F)
transitions, respectively, based on the spectral features observed for other octahedral high-spin cobalt(II) complexes.27–30The energies
of these transitions were used to perform a ligand-field analysis of the trimetaphosphate anion (SectionS7). The ∆o value of 9,280
cm−1obtained from these studies indicates the field strength expe-rienced by a cobalt(II) ion within an Ohcoordination environment
composed of trimetaphosphate anions is comparable to that exerted by water (∆o[Co(H2O)6]2+= 9,200 cm−1).28,29
The31P NMR spectrum of complex 2, recorded in acetonitrile,
features a broad (∆ν1/2= 3100 Hz,Figure S8) singlet located at δ
1680 ppm that is assigned to the three equivalent phosphorus atoms of each [P3O9]3− ligand. Complex 2 features absorptions located
at λmax449 (ε = 10 M−1cm−1), and 666 (ε = 4 M−1cm−1) nm in
the visible region of the electromagnetic spectrum that are assigned to the3T1g(P)←3T1gand3T2g←3T1gtransitions, respectively (
Fig-ure S13). These bands compare well with the absorptions observed for similar octahedral vanadium(III) complexes supported by weak-field ligands.31 The magnetic moment for 2 (µ
eff = 2.99 µB), as
measured by the Evans method26 (CD3CN, 25◦C), is consistent
with a triplet d2electronic configuration.
The [PPN]4[1]·2MeCN and [PPN]3[2]·DME salts have
electro-chemical and solubility properties that are attractive for NARFB applications. The solubilities of both salts in MeCN were de-termined using UV-vis spectroscopy. Acetonitrile was the opti-mal polar-organic solvent choice for the [PPN]4[1]·2MeCN and
[PPN]3[2]·DME electrolytes as it is a commonly used solvent
for NARFB applications on the basis of its wide electrochem-ical window (ca. 5 V) and low viscosity.32 The solubilites of [PPN]4[1]·2MeCN and [PPN]3[2]·DME in MeCN were determined
to be 1.09 ± 0.02 and 0.77 ± 0.06 M, respectively. These solubil-ities compare favorably with those of other transition-metal
com-Cycle Number50 60 70 80 90 100 40 30 20 10 0 Co ulo mb ic Effic ie ncy (%) 100 70 20 0 90 80 60 50 40 30 10 4 5 6 7 8Cycle Number9 10 11 12 13 14 15 Time (d) 3.0 3.5 4.0 2.5 2.0 1.5 P ote nt ial (V ) 1.0 0.5 3.0 2.5 2.0 1.5
Figure 2. Top: Representative voltage vs. time profile recorded between the 4thand 15thcycle of the 1(+)/2(–) cell (5 mM in MeCN, see SectionS8.1 for more detail). Bottom: Coulombic efficiency values of all 100 cycles.
plex electrolytes,13 and exceed the operational concentration of V(acac)3(ca. 50 mM)33by a factor of approximately twenty. The
lower solubility of [PPN]3[2]·DME when compared with that of
[PPN]4[1]·2MeCN can be explained by the fact that complex 2 is
trianionic and therefore is accompanied by three lipophilic PPN+ cations when compared with the four associated with 1.
In MeCN, 1 features a reversible oxidation located at E1/2 = +0.550 V vs. Fc/Fc+ that is assigned to the CoII/IIIcouple, and
2 exhibits a reversible reduction at E1/2= −1.865 V vs. Fc/Fc+
at-tributed to the VIII/IIcouple (Figure 1, A). Based on their combined reversible redox events, complexes 1 and 2 were paired to construct a cell with a theoretical voltage of 2.4 V in which 1 serves as the catholyte and cycles between Co(II) and Co(III) oxidation states, and 2 serves as the anolyte, cycling between V(III) and V(II) ox-idation states. Within this system, both metal complexes undergo one electron redox events, and alternate between tri- and tetraan-ionic species (Figure 1, B).
Such a two-compartment static H-cell was assembled to assess the charge-discharge characteristics of 1 and 2 in MeCN (Section
S8). The H-cell is commonly used to approximate the conditions of a flow system and allows for the use of small volumes of elec-trolyte solutions.13The catholyte and anolyte compartments of the cell contained 5 mM MeCN solutions of 1 and 2, respectively, and Teflon coated stirbars were used in each reservoir to effect efficient mixing. The cell was run with [TBA][PF6] supporting electrolyte,
and the two compartments were separated with a Neosepta® (AS-TOM, Japan) membrane (Figure 1, B). The Neosepta® membrane is an anion-exchange membrane with monoanion permselectiv-ity,34 allowing for the transport of [PF6]− anions with minimal
permeation of the redox-active multiply charged metal metaphos-phate species, [Co(P3O9)2]4−/3−, and [V(P3O9)2]3−/4−.
Elec-trodes constructed of high-surface-area carbon (with 1 cm2active
areas) were used in each compartment of the cell. Galvanostatic cycling of the cell was performed at 80% state of charge (SOC), and the cell was charged and discharged at currents of 0.210 mA (C/2.5) and 0.0525 mA (C/10), respectively (Figure 2, top). Volt-age cutoffs (0.5–3.1 V vs. Ag/Ag+) were selected on the basis of the cyclic voltammograms of each complex to ensure that only the desired redox couples were accessed.
The resulting cell displayed a Vcellof 2.4 V, which is consistent
with the 2.4 V theoretical maximum voltage of the cell based on the CVs of each complex (Figure 1, A). The observed operating cell voltage of 2.4 V is among the highest we are aware of for the state-of-the-art NARFBs involving metal-based redox systems that exhibit high cycling stability.3,9,13,16–18,35,36The charge-discharge cycles of the 1-2 cell have high Coulombic efficiencies exceeding 90% for up to 100 cycles (Figure 2, bottom), with minor losses likely due to degradation of the anolyte after prolonged cycling, and small amounts of species crossover through the membrane (Section
S8.1.1) These efficiencies compare favorably with those reported for other NARFB candidate systems,10,11,35 and are higher than those reported for V(acac)3 in similar cell designs (ca. 70%).8,9
The excellent performance of metal metaphosphate polyanions in NARFB applications is likely due in part to the high chemical and electrochemical stability of the all-inorganic trimetaphosphate ligands. Additionally, the polyanionic nature of all redox-active species relevant to cycling of the present cell is likely a key fac-tor in the effective utilization of existing membrane technology as it contributes to minimal cross-contamination between the anolyte and catholyte compartments. We have addressed the stability of the anolyte and catholyte species within the context of the H-cell configuration and half-cell cyclability studies (SectionS9). Future work will be needed at the full cell level to assess stability bench-marks relevant to an operating flow cell.
75 °C 0 °C 0 50 100 150 200 250 300 350 400 450 350 400 450 500 550 600 650 700 30000 25000 21000 18000 16000 14000 12000 ε (M –1 c m –1) Wavelength (nm) Wavenumber (cm–1) ↑
Figure 3. Variable temperature UV-vis spectra of [PPN]3[Co(P3O9)2]·MeCN (MeCN, 4 mM, –10-75◦C).
It is noted that 2 also has an accessible and reversible VIII/IV couple in MeCN located at E1/2 = +0.869 V vs. Fc/Fc+ (
Fig-ure S12). A symmetric, all-vanadium redox-flow cell with 2 as
both the positive and negative electrodes would surpass the voltage of the current 1-2 system with a total Vcellof 2.7 V. Initial studies
of this all-vanadium cell were met with little success. When the cell was charged, the putative [VIV(P3O9)2]2−complex was generated;
this species is significantly less soluble in MeCN than the starting complex 2, a circumstance that results in precipitation of the active material and failure of the cell. Current efforts are focused on varia-tion of the countercavaria-tion, supporting electrolyte, and/or solvent that may provide adequate solubility enhancements allowing access to the higher voltage all-vanadium cell.
Preliminary testing at the full cell level was conducted, and the resulting redox-flow battery displayed a stable charge-discharge profile for 26 total cycles (10 mM in MeCN, C/5 charge and dis-charge rates, 20% SOC, SectionS10). These promising prelimi-nary results provide support for the use of the 1-2 system in redox-flow battery applications. Future work will focus on more extensive studies at the full cell level and optimization of flow-cell conditions.
Efforts were made to isolate the [CoIII(P
3O9)]3−complex to gain
more insight into the properties of the active catholyte species in all charge states during cell cycling. Proceeding accordingly, com-plex 1 was treated with the tris(p-bromophenyl)aminium radical cation ([(p-BrC6H4)3N]+, Eo= 0.67 V vs. Fc/Fc+, MeCN,
Equa-tion2),37 and the oxidized product ([PPN]3[Co(P3O9)2]·MeCN,
[PPN]3[3]·MeCN) was isolated as a pale yellow solid after
precip-itation from the reaction mixture. The same oxidant has been used to prepare closely related cobalt(III) complexes supported by the Kläui ligand ([(C5H5)Co(P(O)R2)3]−, R = OMe, OEt, OiPr) that
display similar electrochemical behavior when compared with that of 3.27,38 Like [P3O9]3−, the tripodal anionic Kläui ligand binds
metal ions with a facial O3donor set,38–40resulting in an
octahe-dral CoO6coordination environment at the cobalt center.
[PPN]4[CoII(P3O9)2]·2MeCN[(p-BrC6MeCNH4)3N][SbCl6]
– (p-BrC6H4)3N – [PPN][SbCl6]
1 3
[PPN]3[CoIII(P3O9)2]·MeCN
(2) Complex 3 is characterized by a λmaxof 396 nm (MeCN, 25◦C,
ε = 190 M−1cm−1,Figure 3), a59Co NMR resonance of δ 8632 ppm (Figure S16, ∆ν1/2= 222 Hz, 25◦C, MeCN), and a31P NMR
shift of δ 2.1 ppm (Figure S14, ∆ν1/2= 62 Hz, MeCN, 25◦C) that
is assigned to the phosphorus atoms of the [P3O9]3−ligand. This 31P NMR resonance is shifted downfield from the region expected
for a low-spin, S = 0 Co(III) d6complex based on the31P NMR signatures observed for diamagnetic trimetaphosphate species.21,23 The downfield shift suggested that the high-spin S = 2 state may
be populated to some extent at 25◦C. Solution magnetic suscep-tibility data of [PPN]3[3]·MeCN at 25◦C (Evans method, µeff =
3.32 µB, CD3CN) further substantiated contribution from the S = 2
high-spin state. These data in combination with the classification of trimetaphosphate as a weak-field ligand suggest that complex 3 ex-hibits a low-spin *) high-spin transition described by the1A1(Oh)
*
)5T2(Oh) states. In fact, many of the cobalt(III) complexes
sup-ported by the Kläui ligand exhibit singlet-quintet spin-crossover be-havior as well.38–40There is a large body of literature describing isoelectronic (3d6) octahedral iron(II) spin-crossover complexes,41 and it has been well documented that the spin-state transition pro-ceeds directly from S = 0 to S = 2, and never through a detectable S = 1 state.38
Collection of the 31P NMR spectrum of 3 at 25, 45, and 75
◦C displayed corresponding downfield shifts in the [P
3O9]3−
reso-nance ranging from 2.1 (25◦C) to 17.9 ppm (75◦C) (Figure S21). The solution magnetic susceptibility values also increased from µeff
3.32 (25◦C) to 3.74 µB(75◦C); these observations are consistent
with thermally-induced spin crossover to a paramagnetic state at elevated temperature.27,42 The spin-crossover behavior is accom-panied by a visually-detectable loss of color at low temperature (< 5◦C), and an intensification of the yellow color at elevated tem-perature (> 50◦C), as shown inFigure 3. These noticeable color changes allowed us to probe the spin-crossover behavior of 3 fur-ther through collection of variable-temperature UV-vis data (–10-75◦C,Figure 3). The spin equilibrium was modeled by using the absorbance intensities at various temperatures and fitting the data to the Boltzmann distribution.27,40,42,43Values of ∆H, and ∆S were extracted from the UV-vis data and determined to be, 56.02 ± 4.77 JK−1mol−1, and 18.49 ± 2.09 kJmol−1, respectively. These data allowed for calculation of the high-spin percentage at any given temperature, where the high-spin S = 2 percentage at 300 K was determined to be 34.0 ± 7.2%.
Electrochemical investigations of the spin-equilibrium phe-nomenon for spin-crossover complexes reveal that these systems feature half-wave potentials that are highly temperature depen-dent.44 The fact that the spin crossover complex 3 is the active catholyte species generated during cell charging offers a unique op-portunity to potentially expand the overall cell voltage (Vcell) of the
present 1-2 system based on operating temperature.
Supporting Information Available: Experimental details and
characterization data for all complexes including crystallographic data
for [PPN]4[Co(P3O9)2]·2MeCN. This material is available free of
charge via the Internet at http://pubs.acs.org.
Acknowledgement This work was supported by the Skoltech Center for Electrochemical Energy Storage through the Skoltech-MIT collaboration program. The Shao-Horn lab at Skoltech-MIT is thanked for assistance with electrochemical cell cycling experiments. Wes-ley Transue is thanked for performing fits for variable temper-ature UV-vis studies and calculation of ∆S and ∆H values of [PPN]3[Co(P3O9)2]·MeCN.
References
(1) (a) Ponce de León, C.; Frías-Ferrer, A.; González-García, J.; Szánto, D.; Walsh, F. J Power Sources 2006, 160, 716–732; (b) Lin, K.; Chen, Q.; Ger-hardt, M. R.; Tong, L.; Kim, S. B.; Eisenach, L.; Valle, A. W.; Hardee, D.; Gordon, R. G.; Aziz, M. J.; Marshak, M. P. Science 2015, 349, 1529– 1532; (c) Huskinson, B.; Marshak, M. P.; Suh, C.; Er, S.; Gerhardt, M. R.; Galvin, C. J.; Chen, X.; Aspuru-Guzik, A.; Gordon, R. G.; Aziz, M. J. Nature2014, 505, 195–198; (d) Bartolozzi, M. J. Power Sources 1989, 27, 219–234; (e) Skyllas-Kazacos, M.; Rychcik, M.; Robins, R. G.; Fane, A. G.; Green, M. A. J. Electrochem. Soc. 1986, 133, 1057–1058; (f) Dunn, B.; Kamath, H.; Tarascon, J.-M. Science 2011, 334, 928–935. (2) (a) Skyllas-Kazacos, M.; Chakrabarti, M. H.; Hajimolana, S. A.;
Mjalli, F. S.; Saleem, M. J. Electrochem. Soc. 2011, 158, R55–R79; (b) Park, M.; Ryu, J.; Wang, W.; Cho, J. Nat. Rev. Mater. 2016, 2, 16080; (c) Huang, Y.; Gu, S.; Yan, Y.; Li, S. F. Y. Curr. Opin. Chem. Eng. 2015, 8,
105–113; (d) Liu, T.; Wei, X.; Nie, Z.; Sprenkle, V.; Wang, W. Adv. Energy Mater.2016, 6, 1501449; (e) Hu, B.; DeBruler, C.; Rhodes, Z.; Liu, T. L. J. Am. Chem. Soc.2017, 139, 1207–1214.
(3) Cappillino, P. J.; Pratt, H. D.; Hudak, N. S.; Tomson, N. C.; Ander-son, T. M.; Anstey, M. R. Adv. Energy Mater. 2014, 4, 1300566. (4) Skyllas-Kazacos, M.; Cao, L.; Kazacos, M.; Kausar, N.; Mousa, A.
Chem-SusChem2016, 9, 1521–1543.
(5) (a) Shin, S.-H.; Yun, S.-H.; Moon, S.-H. RSC Adv. 2013, 3, 9095–9116; (b) Zhao, Y.; Ding, Y.; Li, Y.; Peng, L.; Byon, H. R.; Goodenough, J. B.; Yu, G. Chem. Soc. Rev. 2015, 44, 7968–7996; (c) Goodenough, J.; Kim, Y. J. Power Sources2011, 196, 6688–6694; (d) Yang, Z.; Zhang, J.; Kintner-Meyer, M. C. W.; Lu, X.; Choi, D.; Lemmon, J. P.; Liu, J. Chem. Rev. 2011, 111, 3577–3613.
(6) (a) Winsberg, J.; Hagemann, T.; Janoschka, T.; Hager, M. D.; Schu-bert, U. S. Angew. Chem. Int. Ed. 2017, 56, 686–711; (b) Duan, W. et al. ACS Energy Lett.2017, 2, 1156–1161; (c) Wei, X.; Xu, W.; Vijayaku-mar, M.; Cosimbescu, L.; Liu, T.; Sprenkle, V.; Wang, W. Adv. Mater. 2014, 26, 7649–7653.
(7) Singh, P. J. Power Sources 1984, 11, 135–142.
(8) Shinkle, A. A.; Sleightholme, A. E.; Griffith, L. D.; Thompson, L. T.; Mon-roe, C. W. J. Power Sources 2012, 206, 490–496.
(9) Liu, Q.; Sleightholme, A. E.; Shinkle, A. A.; Li, Y.; Thompson, L. T. Electrochem. Commun.2009, 11, 2312–2315.
(10) Liu, Q.; Shinkle, A. A.; Li, Y.; Monroe, C. W.; Thompson, L. T.; Sleightholme, A. E. Electrochem. Commun. 2010, 12, 1634–1637. (11) Suttil, J. A.; Kucharyson, J. F.; Escalante-Garcia, I. L.; Cabrera, P. J.;
James, B. R.; Savinell, R. F.; Sanford, M. S.; Thompson, L. T. J. Mater. Chem. A2015, 3, 7929–7938.
(12) Cabrera, P. J.; Yang, X.; Suttil, J. A.; Hawthorne, K. L.; Brooner, R. E. M.; Sanford, M. S.; Thompson, L. T. J. Phys. Chem. C 2015, 119, 15882– 15889.
(13) Sevov, C. S.; Fisher, S. L.; Thompson, L. T.; Sanford, M. S. J. Am. Chem. Soc.2016, 138, 15378–15384.
(14) Eisenberg, R.; Gray, H. B. Inorg. Chem. 2011, 50, 9741–9751.
(15) (a) Hwang, B.; Park, M.-S.; Kim, K. ChemSusChem 2015, 8, 310–314; (b) Zhao, Y.; Ding, Y.; Song, J.; Li, G.; Dong, G.; Goodenough, J. B.; Yu, G. Angew. Chem. Int. Ed.2014, 53, 11036–11040; (c) Ding, Y.; Zhao, Y.; Li, Y.; Goodenough, J. B.; Yu, G. Energy Environ. Sci. 2017, 10, 491–497. (16) Cabrera, P. J.; Yang, X.; Suttil, J. A.; Brooner, R. E. M.; Thompson, L. T.;
Sanford, M. S. Inorg. Chem. 2015, 54, 10214–10223.
(17) Mun, J.; Lee, M.-J.; Park, J.-W.; Oh, D.-J.; Lee, D.-Y.; Doo, S.-G. Elec-trochem. Solid-State Lett.2012, 15, A80–A82.
(18) Gong, K.; Fang, Q.; Gu, S.; Li, S. F. Y.; Yan, Y. Energy Environ. Sci. 2015, 8, 3515–3530.
(19) Nawi, M. A.; Riechel, T. L. Inorg. Chem. 1981, 20, 1974–1978. (20) Troutman, M.; Ravichandran, R.; Zingg, J.; King, R. Flame Retardant
Coatings, 2004, US Patent App. 10/406,484.
(21) Stauber, J. M.; Cummins, C. C. Inorg. Chem. 2017, 56, 3022–3029. (22) Jiang, Y.; Chakarawet, K.; Kohout, A. L.; Nava, M.; Marino, N.;
Cum-mins, C. C. J. Am. Chem. Soc. 2014, 136, 11894–11897.
(23) Chakarawet, K.; Knopf, I.; Nava, M.; Jiang, Y.; Stauber, J. M.; Cum-mins, C. C. Inorg. Chem. 2016, 55, 6178–6185.
(24) Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456–463.
(25) Suh, I.-H.; Lee, J.-H.; Lim, S.-S.; Park, J.-R.; Choo, G. H.; Kwak, S.-H.; Do, Y.; Lee, H. J. Chungnam J. Sci 1994, 21, 124.
(26) Evans, D. F. J. Chem. Soc. 1959, 2003–2005.
(27) Kläui, W.; Eberspach, W.; Guetlich, P. Inorg. Chem. 1987, 26, 3977–3982. (28) Chemistry of the Elements, 2nd ed.; Greenwood, N., , Earnshaw, A., Eds.;
Butterworth-Heinemann: Oxford, 1997; pp 1113–1143. (29) Jørgensen, C. K. Acta Chem. Scand. 1954, 8, 1495–1501.
(30) (a) Pappalardo, R. Philos. Mag. 1959, 4, 219–242; (b) Holmes, O. G.; Mc-Clure, D. S. J. Chem. Phys. 1957, 26, 1686–1694; (c) Askalani, P. Transi-tion Met. Chem.1986, 11, 469–471.
(31) (a) Tregenna-Piggott, P. L. W.; Spichiger, D.; Carver, G.; Frey, B.; Meier, R.; Weihe, H.; Cowan, J. A.; McIntyre, G. J.; Zahn, G.; Barra, A.-L. Inorg. Chem.2004, 43, 8049–8060; (b) Kittilstved, K. R.; Sorgho, L. A.; Amstutz, N.; Tregenna-Piggott, P. L.; Hauser, A. Inorg. Chem. 2009, 48, 7750–7764; (c) Machin, D. J.; Murray, K. S. J. Chem. Soc. A 1967, 1498– 1504.
(32) Barrosse-Antle, L.; Bond, A.; Compton, R.; O’Mahony, A.; Rogers, E.; Silvester, D. Chem. Asian J. 2010, 5, 202–230.
(33) Pan, F.; Wang, Q. Molecules 2015, 20, 20499–20517.
(34) (a) Długołe¸cki, P.; Anet, B.; Metz, S. J.; Nijmeijer, K.; Wessling, M. J. Membr. Sci.2010, 346, 163–171; (b) http://www.astom-corp.jp. (35) Sleightholme, A. E.; Shinkle, A. A.; Liu, Q.; Li, Y.; Monroe, C. W.;
Thompson, L. T. J. Power Sources 2011, 196, 5742–5745.
(36) (a) Chakrabarti, M.; Dryfe, R.; Roberts, E. Electrochim Acta 2007, 52, 2189–2195; (b) Yamamura, T.; Shiokawa, Y.; Yamana, H.; Moriyama, H. Electrochim Acta2002, 48, 43–50; (c) Kim, J.-H.; Kim, K. J.; Park, M.-S.; Lee, N. J.; Hwang, U.; Kim, H.; Kim, Y.-J. Electrochem. Commun. 2011, 13, 997–1000; (d) Zhang, D.; Lan, H.; Li, Y. J. Power Sources 2012, 217, 199–203; (e) Pratt, H. D.; Hudak, N. S.; Fang, X.; Anderson, T. M. J. of Power Sources2013, 236, 259–264; (f) Wang, W.; Luo, Q.; Li, B.; Wei, X.; Li, L.; Yang, Z. Adv. Funct. Mater. 2013, 23, 970–986.
(37) (a) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877–910; (b) Bell, F. A.; Ledwith, A.; Sherrington, D. C. J. Chem. Soc. C 1969, 2719– 2720.
(38) Kläui, W. J. Chem. Soc., Chem. Commun. 1979, 700.
(39) Kläui, W.; Otto, H.; Eberspach, W.; Buchholz, E. Ber. 1982, 115, 1922– 1933.
(40) Navon, G.; Kläui, W. Inorg. Chem. 1984, 23, 2722–2725.
(41) (a) Gütlich, P.; Hauser, A.; Spiering, H. Angew. Chem. Int. Ed. 1994, 33, 2024–2054; (b) König, E. Coord. Chem. Rev. 1968, 3, 471–495; (c) Goodwin, H. A. Coord. Chem. Rev. 1976, 18, 293–325; (d) Gütlich, P.;
Hauser, A. Coord. Chem. Rev. 1990, 97, 1–22; (e) Gutlich, P.; Garcia, Y.; Goodwin, H. A. Chem. Soc. Rev. 2000, 29, 419–427.
(42) King, E. R.; Sazama, G. T.; Betley, T. A. J. Am. Chem. Soc. 2012, 134, 17858–17861.
(43) Creutz, S. E.; Peters, J. C. Inorg. Chem. 2016, 55, 3894–3906.
(44) (a) Kadish, K. M.; Das, K.; Schaeper, D.; Merrill, C. L.; Welch, B. R.; Wil-son, L. J. Inorg. Chem. 1980, 19, 2816–2821; (b) Kadish, K. M.; Su, C. H.; Wilson, L. J. Inorg. Chem. 1982, 21, 2312–2314; (c) Zhu, T.; Su, C. H.; Schaeper, D.; Lemke, B. K.; Wilson, L. J.; Kadish, K. M. Inorg. Chem. 1984, 23, 4345–4349; (d) Turner, J. W.; Schultz, F. A. Coord. Chem. Rev. 2001, 219, 81–97.
For Table of Contents Entry Only Potential (V vs. Fc/Fc+) 1.0 0.5 0 -1.5 -2.0 -2.5 4–/3– O P O O O O O O O O P O P CoII/III P O O O O O O O O P P P O O O O O O O O P O P VIII/II P O O O O O O O O O P P 3–/4–
