Cobalt and Vanadium Trimetaphosphate Polyanions: Synthesis

Dec 12, 2017 - Friedl, Lebedeva, Porfyrakis, Stimming, and Chamberlain. 2018 140 (1), pp 401–405. Abstract: Redox flow batteries have the potential ...
1 downloads 0 Views 2MB Size
Subscriber access provided by READING UNIV

Communication

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 J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/jacs.7b08751 • Publication Date (Web): 12 Dec 2017 Downloaded from http://pubs.acs.org on December 12, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Journal of the American Chemical Society is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

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

Journal of the American Chemical Society

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: [email protected]; [email protected]

ABSTRACT: An electrochemical cell consisting of cobalt ([CoII/III (P3 O9 )2 ]4−/3− ) and vanadium ([VIII/II (P3 O9 )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 cycles. The [Co(P3 O9 )2 ]4− (1) and [V(P3 O9 )2 ]3− (2) complexes have favorable properties for flow-battery applications that include reversible redox chemistry, high stability toward electrochemical 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 characterized by 31 P 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-BrC6 H4 )3 N]+ allowed for isolation of the one-electron oxidized spin-crossover complex, [Co(P3 O9 )2 ]3− (3), which is the active catholyte species generated during cell charging. The presence of a spin-crossover complex is a feature of the present system that may enable voltage window variation as a function of temperature. 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 gridscale energy storage. 1 RFBs 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 current 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 disadvantages that include low energy density (∼50 Whl−1 ), limited operational temperatures (0–100 ◦ C), and a narrow electrochemical window (1.2 V). 1,5 To 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 electrolytes supported by organic ligands (dithiolate, 3,14 cyclopentadienyl, 15 bipyridine, 16,17 acetylacetonate 9,11 ) suffer from numerous drawbacks such as low cyclability, 18 high production cost, † Massachusetts ‡ Skolkovo

Institute of Technology Institute of Science and Technology

high flammability, limited electrochemical stability, 8,18 ligand dissociation, 8,13,19 and cross-contamination between the anolyte and catholyte compartments; 10 these drawbacks have thus far prevented wide-scale adoption of NARFBs. Herein, we report a new approach to NARFB electrolyte design through the use of redox-active cyclic phosphate metal complexes. This work presents trimetaphosphate ([P3 O9 ]3− ) anions as all-inorganic supporting ligands for cobalt and vanadium ions (Figure 1, B; [PPN]4 [Co(P3 O9 )2 ]·2MeCN, [PPN]4 [1]·2MeCN; [PPN]3 [V(P3 O9 )2 ]·DME, [PPN]3 [2]·DME), in which the redoxactive metal center’s coordination sphere is completed entirely by a pair of metaphosphate rings. The electrochemical inertness of the redox inactive cyclic phosphates ([PO3 ]n n− ), coupled with their low cost, low toxicity, and flame-retardant nature 20 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 trimetaphosphate ring aids in disfavoring ligand dissociation, 13 and 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 lifetimes of other NARFB candidate systems. 9,11 The transition-metal complexes are also polyanionic in all charge states that are relevant to cycling of the RFB, a property that minimizes crossover of active species through membrane separators. Additionally, the solubilities 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] + Mn+(acac)n + n–2 H

–n acacH

[PPN]y[Mn+(P3O9)2] 1, M = Co, n = 2, y = 4 2, M = V, n = 3, y = 3

(1) We have previously shown that monohydrogen trimetaphosphate ([P3 O9 H]2− ) is capable of effecting ligand exchange through protonolyis of acetylacetonate (acac) from [OTi(acac)2 ]2 to generate the terminal titanyl trimetaphosphate complex, [OTiP3 O9 (acac)]2− . 21 Employing this strategy of protolytic ligand replacement, salts of complexes [Co(P3 O9 )2 ]4− (1) and [V(P3 O9 )2 ]3− (2) were prepared through treatment of [PPN]2 [P3 O9 H] (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 transitionmetal metaphosphate complexes, making the preparation of the PPN+ salts of 1 and 2 convenient and straightforward. Preparation of [PPN]3 [V(P3 O9 )2 ]·DME, however, requires three equivalents

ACS Paragon Plus Environment 1

+

Journal of the American Chemical Society

A

B

2 μA

e– charging

+



ΔE = 2.4 V porous carbon electrode O O P O

O

O

O

O

O

-2.5

O

-2.0

P

O

-1.5

PF6–

1

O O P O

O

O

O

-0.5 -1.0 Potential (V vs. Fc/Fc+)

P

O

P

P O

O

0

O

CoII/III O

0.5

P

4–/3– O

O

1.0

P

O

O

P

3–/4– O

O

VIII/II

P

O

O

O

[V (P3O9)2]

P

4–

O

O II

P

O

[VIII(P3O9)2]3–

O

[CoII(P3O9)2]4–

O

[CoIII(P3O9)2]3–

O

2 mono-anion selective membrane

e– discharging

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(P3 O9 )2 ]4−/3− (catholyte, left), and [V(P3 O9 )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. Figure 1.

1680 ppm that is assigned to the three equivalent phosphorus atoms of each [P3 O9 ]3− ligand. Complex 2 features absorptions located at λ max 449 (ε = 10 M−1 cm−1 ), and 666 (ε = 4 M−1 cm−1 ) nm in the visible region of the electromagnetic spectrum that are assigned to the 3 T1g (P)←3 T1g and 3 T2g ←3 T1g transitions, respectively (Figure S13). These bands compare well with the absorptions observed for similar octahedral vanadium(III) complexes supported by weakfield ligands. 31 The magnetic moment for 2 (µ eff = 2.99 µ B ), as measured by the Evans method 26 (CD3 CN, 25 ◦ C), is consistent with a triplet d 2 electronic configuration. The [PPN]4 [1]·2MeCN and [PPN]3 [2]·DME salts have electrochemical and solubility properties that are attractive for NARFB applications. The solubilities of both salts in MeCN were determined using UV-vis spectroscopy. Acetonitrile was the optimal 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 electrochemical 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 solubilities compare favorably with those of other transition-metal com4

5

8

7

Cycle Number 9 10

11

12

13

14

15

2.5 2.0 1.5 1.0 0.5 1.5 100 90 80 70 60 50 40 30 20 10 0

0

10

2.0

20

2.5 Time (d)

30

3.0

40 50 60 Cycle Number

3.5

70

80

4.0

90

100

Top: Representative voltage vs. time profile recorded between the 4th and 15th cycle of the 1(+)/2(–) cell (5 mM in MeCN, see Section S8.1 for more detail). Bottom: Coulombic efficiency values of all 100 cycles. Figure 2.

ACS Paragon Plus Environment 2

6

3.0 Potential (V)

of acid for protonation of all three acac ligands of V(acac)3 . Two equivalents are provided by [PPN]2 [P3 O9 H], and the third is provided by addition of one equivalent of p-CF3 C6 H4 COOH. The p-CF3 C6 H4 COOH reagent was chosen as an appropriate acid for protonation of acac from V(acac)3 based on pKa values, 24 and due to the formation of the [PPN][p-CF3 C6 H4 COO] byproduct, which is easily separated away from [PPN]3 [2]·DME by thorough washing 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) complex was confirmed by an X-ray diffraction study (Section S12). The solid-state structure of 1 displays the cobalt(II) ion within an all-oxygen octahedral coordination environment provided by the two [P3 O9 ]3− rings with a Co–Oavg distance of 2.108(9) Å (Figure 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 reported previously. 25 Complex 1 is characterized by a 31 P NMR resonance located at δ 616.7 ppm (∆ν 1/2 = 466 Hz, MeCN, 25 ◦ C, Figure S1) that is assigned to the phosphorus atoms of the [P3 O9 ]3− ligand. This signal is shifted significantly downfield from that of monohydrogen trimetaphosphate (δ –22.5 ppm, MeCN) 23 due to the influence imparted by the paramagnetic Co(II) d 7 metal center. Consistent with the presence of a high-spin Co(II) ion, complex 1 displays an S = 3/2 electronic configuration at 25 ◦ C in CD3 CN, as determined by the Evans method (µeff = 4.10 µB ). 26 The electronic spectrum of 1 measured in MeCN (Figure S6, Figure S7) is characterized by absorptions located at λ max 487 (sh), 540 (ε = 5 M−1 cm−1 ), and 1135 (ε = 1 M−1 cm−1 ), which are assigned to the 4 A2g (F)←4 T1g (F), 4 T1g (P)←4 T1g (F), and 4 T2g (F)←4 T1g (F) transitions, respectively, based on the spectral features observed for other octahedral high-spin cobalt(II) complexes. 27–30 The energies of these transitions were used to perform a ligand-field analysis of the trimetaphosphate anion (Section S7). The ∆o value of 9,280 cm−1 obtained from these studies indicates the field strength experienced by a cobalt(II) ion within an Oh coordination environment composed of trimetaphosphate anions is comparable to that exerted by water (∆o [Co(H2 O)6 ]2+ = 9,200 cm−1 ). 28,29 The 31 P NMR spectrum of complex 2, recorded in acetonitrile, features a broad (∆ν 1/2 = 3100 Hz, Figure S8) singlet located at δ

Coulombic Efficiency (%)

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

Page 2 of 12

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

Journal of the American Chemical Society

plex electrolytes, 13 and exceed the operational concentration of V(acac)3 (ca. 50 mM) 33 by 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/III couple, and 2 exhibits a reversible reduction at E1/2 = −1.865 V vs. Fc/Fc+ attributed to the VIII/II couple (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) oxidation states. Within this system, both metal complexes undergo one electron redox events, and alternate between tri- and tetraanionic 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 electrolyte solutions. 13 The 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® (ASTOM, Japan) membrane (Figure 1, B). The Neosepta® membrane is an anion-exchange membrane with monoanion permselectivity, 34 allowing for the transport of [PF6 ]− anions with minimal permeation of the redox-active multiply charged metal metaphosphate species, [Co(P3 O9 )2 ]4−/3− , and [V(P3 O9 )2 ]3−/4− . Electrodes constructed of high-surface-area carbon (with 1 cm2 active 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). Voltage 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 V cell of 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. 3,9,13,16–18,35,36 The 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 factor 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 (Section S9). Future work will be needed at the full cell level to assess stability benchmarks relevant to an operating flow cell.

Wavenumber (cm–1) 30000

21000

18000

16000

14000

12000

400 350



300 250 200 150

0 °C

75 °C

100 50 0 350

400

450

500

550

600

650

700

Wavelength (nm)

3. Variable temperature UV-vis [PPN]3 [Co(P3 O9 )2 ]·MeCN (MeCN, 4 mM, –10-75 ◦ C). Figure

spectra

of

It is noted that 2 also has an accessible and reversible VIII/IV couple in MeCN located at E 1/2 = +0.869 V vs. Fc/Fc+ (Figure 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 V cell of 2.7 V. Initial studies of this all-vanadium cell were met with little success. When the cell was charged, the putative [VIV (P3 O9 )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 variation of the countercation, 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 discharge rates, 20% SOC, Section S10). These promising preliminary results provide support for the use of the 1-2 system in redoxflow 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 (P3 O9 )]3− complex to gain more insight into the properties of the active catholyte species in all charge states during cell cycling. Proceeding accordingly, complex 1 was treated with the tris(p-bromophenyl)aminium radical cation ([(p-BrC6 H4 )3 N]+ , E o = 0.67 V vs. Fc/Fc+ , MeCN, Equation 2), 37 and the oxidized product ([PPN]3 [Co(P3 O9 )2 ]·MeCN, [PPN]3 [3]·MeCN) was isolated as a pale yellow solid after precipitation from the reaction mixture. The same oxidant has been used to prepare closely related cobalt(III) complexes supported by the Kläui ligand ([(C5 H5 )Co(P(O)R2 )3 ]− , R = OMe, OEt, Oi Pr) that display similar electrochemical behavior when compared with that of 3. 27,38 Like [P3 O9 ]3− , the tripodal anionic Kläui ligand binds metal ions with a facial O3 donor set, 38–40 resulting in an octahedral CoO6 coordination environment at the cobalt center. [PPN]4[CoII(P3O9)2]·2MeCN

1

[(p-BrC6H4)3N][SbCl6] MeCN – (p-BrC6H4)3N – [PPN][SbCl6]

[PPN]3[CoIII(P3O9)2]·MeCN

3

(2) Complex 3 is characterized by a λ max of 396 nm (MeCN, 25 ◦ C, ε = 190 M−1 cm−1 , Figure 3), a 59 Co NMR resonance of δ 8632 ppm (Figure S16, ∆ν 1/2 = 222 Hz, 25 ◦ C, MeCN), and a 31 P NMR shift of δ 2.1 ppm (Figure S14, ∆ν 1/2 = 62 Hz, MeCN, 25 ◦ C) that is assigned to the phosphorus atoms of the [P3 O9 ]3− ligand. This 31 P NMR resonance is shifted downfield from the region expected for a low-spin, S = 0 Co(III) d 6 complex based on the 31 P NMR signatures observed for diamagnetic trimetaphosphate species. 21,23 The downfield shift suggested that the high-spin S = 2 state may

ACS Paragon Plus Environment 3

25000

450

ε (M–1 cm–1)

Page 3 of 12

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

be populated to some extent at 25 ◦ C. Solution magnetic susceptibility data of [PPN]3 [3]·MeCN at 25 ◦ C (Evans method, µ eff = 3.32 µ B , CD3 CN) 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 exhibits a low-spin * ) high-spin transition described by the 1 A1 (Oh ) 5 * ) T2 (Oh ) states. In fact, many of the cobalt(III) complexes supported by the Kläui ligand exhibit singlet-quintet spin-crossover behavior as well. 38–40 There is a large body of literature describing isoelectronic (3d 6 ) octahedral iron(II) spin-crossover complexes, 41 and it has been well documented that the spin-state transition proceeds directly from S = 0 to S = 2, and never through a detectable S = 1 state. 38 Collection of the 31 P NMR spectrum of 3 at 25, 45, and 75 ◦ C displayed corresponding downfield shifts in the [P O ]3− reso3 9 nance ranging from 2.1 (25 ◦ C) to 17.9 ppm (75 ◦ C) (Figure S21). The solution magnetic susceptibility values also increased from µ eff 3.32 (25 ◦ C) to 3.74 µ B (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 accompanied by a visually-detectable loss of color at low temperature (< 5 ◦ C), and an intensification of the yellow color at elevated temperature (> 50 ◦ C), as shown in Figure 3. These noticeable color changes allowed us to probe the spin-crossover behavior of 3 further through collection of variable-temperature UV-vis data (–1075 ◦ C, Figure 3). The spin equilibrium was modeled by using the absorbance intensities at various temperatures and fitting the data to the Boltzmann distribution. 27,40,42,43 Values of ∆H, and ∆S were extracted from the UV-vis data and determined to be, 56.02 ± 4.77 JK−1 mol−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 phenomenon for spin-crossover complexes reveal that these systems feature half-wave potentials that are highly temperature dependent. 44 The fact that the spin crossover complex 3 is the active catholyte species generated during cell charging offers a unique opportunity to potentially expand the overall cell voltage (V cell ) 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(P3 O9 )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 SkoltechMIT collaboration program. The Shao-Horn lab at MIT is thanked for assistance with electrochemical cell cycling experiments. Wesley Transue is thanked for performing fits for variable temperature UV-vis studies and calculation of ∆S and ∆H values of [PPN]3 [Co(P3 O9 )2 ]·MeCN.

(3) (4) (5)

(6)

(7) (8) (9) (10) (11) (12) (13) (14) (15)

(16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31)

(32) (33) (34) (35) (36)

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.; Gerhardt, 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. Nature 2014, 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,

(37) (38) (39) (40) (41)

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. Cappillino, P. J.; Pratt, H. D.; Hudak, N. S.; Tomson, N. C.; Anderson, T. M.; Anstey, M. R. Adv. Energy Mater. 2014, 4, 1300566. Skyllas-Kazacos, M.; Cao, L.; Kazacos, M.; Kausar, N.; Mousa, A. ChemSusChem 2016, 9, 1521–1543. (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 Sources 2011, 196, 6688–6694; (d) Yang, Z.; Zhang, J.; KintnerMeyer, M. C. W.; Lu, X.; Choi, D.; Lemmon, J. P.; Liu, J. Chem. Rev. 2011, 111, 3577–3613. (a) Winsberg, J.; Hagemann, T.; Janoschka, T.; Hager, M. D.; Schubert, 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.; Vijayakumar, M.; Cosimbescu, L.; Liu, T.; Sprenkle, V.; Wang, W. Adv. Mater. 2014, 26, 7649–7653. Singh, P. J. Power Sources 1984, 11, 135–142. Shinkle, A. A.; Sleightholme, A. E.; Griffith, L. D.; Thompson, L. T.; Monroe, C. W. J. Power Sources 2012, 206, 490–496. Liu, Q.; Sleightholme, A. E.; Shinkle, A. A.; Li, Y.; Thompson, L. T. Electrochem. Commun. 2009, 11, 2312–2315. Liu, Q.; Shinkle, A. A.; Li, Y.; Monroe, C. W.; Thompson, L. T.; Sleightholme, A. E. Electrochem. Commun. 2010, 12, 1634–1637. 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. A 2015, 3, 7929–7938. 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. Sevov, C. S.; Fisher, S. L.; Thompson, L. T.; Sanford, M. S. J. Am. Chem. Soc. 2016, 138, 15378–15384. Eisenberg, R.; Gray, H. B. Inorg. Chem. 2011, 50, 9741–9751. (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. Cabrera, P. J.; Yang, X.; Suttil, J. A.; Brooner, R. E. M.; Thompson, L. T.; Sanford, M. S. Inorg. Chem. 2015, 54, 10214–10223. Mun, J.; Lee, M.-J.; Park, J.-W.; Oh, D.-J.; Lee, D.-Y.; Doo, S.-G. Electrochem. Solid-State Lett. 2012, 15, A80–A82. Gong, K.; Fang, Q.; Gu, S.; Li, S. F. Y.; Yan, Y. Energy Environ. Sci. 2015, 8, 3515–3530. Nawi, M. A.; Riechel, T. L. Inorg. Chem. 1981, 20, 1974–1978. Troutman, M.; Ravichandran, R.; Zingg, J.; King, R. Flame Retardant Coatings, 2004, US Patent App. 10/406,484. Stauber, J. M.; Cummins, C. C. Inorg. Chem. 2017, 56, 3022–3029. Jiang, Y.; Chakarawet, K.; Kohout, A. L.; Nava, M.; Marino, N.; Cummins, C. C. J. Am. Chem. Soc. 2014, 136, 11894–11897. Chakarawet, K.; Knopf, I.; Nava, M.; Jiang, Y.; Stauber, J. M.; Cummins, C. C. Inorg. Chem. 2016, 55, 6178–6185. Bordwell, F. G. Acc. Chem. Res. 1988, 21, 456–463. 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. Evans, D. F. J. Chem. Soc. 1959, 2003–2005. Kläui, W.; Eberspach, W.; Guetlich, P. Inorg. Chem. 1987, 26, 3977–3982. Chemistry of the Elements, 2nd ed.; Greenwood, N., , Earnshaw, A., Eds.; Butterworth-Heinemann: Oxford, 1997; pp 1113–1143. Jørgensen, C. K. Acta Chem. Scand. 1954, 8, 1495–1501. (a) Pappalardo, R. Philos. Mag. 1959, 4, 219–242; (b) Holmes, O. G.; McClure, D. S. J. Chem. Phys. 1957, 26, 1686–1694; (c) Askalani, P. Transition Met. Chem. 1986, 11, 469–471. (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. Barrosse-Antle, L.; Bond, A.; Compton, R.; O’Mahony, A.; Rogers, E.; Silvester, D. Chem. Asian J. 2010, 5, 202–230. Pan, F.; Wang, Q. Molecules 2015, 20, 20499–20517. (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. Sleightholme, A. E.; Shinkle, A. A.; Liu, Q.; Li, Y.; Monroe, C. W.; Thompson, L. T. J. Power Sources 2011, 196, 5742–5745. (a) Chakrabarti, M.; Dryfe, R.; Roberts, E. Electrochim Acta 2007, 52, 2189–2195; (b) Yamamura, T.; Shiokawa, Y.; Yamana, H.; Moriyama, H. Electrochim Acta 2002, 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 Sources 2013, 236, 259–264; (f) Wang, W.; Luo, Q.; Li, B.; Wei, X.; Li, L.; Yang, Z. Adv. Funct. Mater. 2013, 23, 970–986. (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. Kläui, W. J. Chem. Soc., Chem. Commun. 1979, 700. Kläui, W.; Otto, H.; Eberspach, W.; Buchholz, E. Ber. 1982, 115, 1922– 1933. Navon, G.; Kläui, W. Inorg. Chem. 1984, 23, 2722–2725. (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.;

ACS Paragon Plus Environment 4

Page 4 of 12

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

Journal of the American Chemical Society

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.; Wilson, 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.

ACS Paragon Plus Environment 5

Journal of the American Chemical Society For Table of Contents Entry Only

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

e–

+

O

P

O

O

O

O

O

O

O

O O

O

O

O

e–

P

O

P

3–/4– O

O

VIII/II

P

O

P

O O P

O

O

CoII/III

P

O

O

P

P

4–/3– O

O

O

O

O

O

O O P

P

O

O

O

P

O

O



ACS Paragon Plus Environment 6

Page 6 of 12

2 [PPN]2[P3O9H] + Mn+(acac)n Page 7 of 12 1 2 3 4 5

Journal of the American Chemical Society

+ n–2 H

+

–n acacH

ACS Paragon Plus Environment

[PPN]y[Mn+(P3O9)2] 1, M = Co, n = 2, y = 4 2, M = V, n = 3, y = 3

A

2 μA

B

e– charging

+

Page 8 of 12



ΔE = 2.4 V porous carbon electrode

-1.5

-2.0

-2.5 ACS Paragon Plus Environment

O

-0.5 -1.0 + Potential (V vs. Fc/Fc )

O

O

0

II/III

O

0.5

P

O

O

O

O

O

1.0

P



P

O

P

O

O

III/II

V

1 e– discharging

P

P

O

Co

PF6

P

O P

O

O

O

P

O

P

O

[V (P3O9)2]

4–

O

O

3–/4–

P

O

II

O P

O

[V (P3O9)2]

3–

O

III

O

O

O

[Co (P3O9)2]

4–

O

4–/3–

O

II

O

O

O

O

[Co (P3O9)2]

3–

O

III

O

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

Journal of the American Chemical Society

2

mono-anion selective membrane

Page 9 of 12

Coulombic Efficiency (%)

Potential (V)

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

4

5

6

Cycle Number 8 9 10 Journal of the American Chemical Society

7

11

12

13

14

15

3.0 2.5 2.0 1.5 1.0 0.5 1.5 100 90 80 70 60 50 40 30 20 10 0

0

10

2.0

20

2.5 Time (d)

30

3.0

40 50 60 Cycle Number ACS Paragon Plus Environment

3.5

70

80

4.0

90

100

Journal of the American Chemical Society 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

ACS Paragon Plus Environment

Page 10 of 12

Wavenumber (cm–1)

Page 11 of 12

Journal of the American Chemical Society

30000

ε (M–1 cm–1)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29

25000

21000

18000

16000

14000

12000

450 400 350



300 250 200 150

0 °C

75 °C

100 50 0 350

400

ACS Paragon Plus 450 500Environment 550

Wavelength (nm)

600

650

700

e–

Journal+ of the American Chemical Page 12 of 12 – Society

O

O O P O

O

O

P

4–/3–

O

O

P

O

O O P O

O

O

CoII/III

O

O

P

O

P

3–/4– O

O

VIII/II O

O

O

P P ACS Paragon Plus Environment P P P O

O

O

O

O

e–

O

O

O

P

O

O

O

O

O

O

1 2 3 4 5 6 7 8 9