Effect of Hydrofluoroether Cosolvent Addition on Li Solvation in

Nov 16, 2016 - Dallek , S.; James , S. D.; Kilroy , W. P. Exothermic Reactions among ..... Angewandte Chemie International Edition 2017 56 (47), 15118...
0 downloads 0 Views 1MB Size
Subscriber access provided by University of Sussex Library

Article

The effect of hydrofluoroether cosolvent addition on Li solvation in acetonitrile-based solvate electrolytes and its influence on S reduction in a Li-S battery Kimberly A. See, Heng-Liang Wu, Kah Chun Lau, Minjeong Shin, Lei Cheng, Mahalingam Balasubramanian, Kevin G. Gallagher, Larry A Curtiss, and Andrew A. Gewirth ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b11358 • Publication Date (Web): 16 Nov 2016 Downloaded from http://pubs.acs.org on November 21, 2016

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.

ACS Applied Materials & Interfaces 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 13

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

ACS Applied Materials & Interfaces

The Effect of Hydrofluoroether Cosolvent Addition on Li solvation in Acetonitrile-Based Solvate Electrolytes and Its Influence on S Reduction in a Li-S Battery Kimberly A. See,#,±,† Heng-Liang Wu,#,±,† Kah Chun Lau,±,ǁ Minjeong Shin,±,† Lei Cheng,±,‡ Mahalingam Balasubramanian,±,∞ Kevin G. Gallagher, ±,§ Larry A. Curtiss,±,‡ and Andrew A. Gewirth*,±,† #

These authors contributed equally to this work.

±

Joint Center for Energy Storage Research, 9700 S. Cass Avenue, Argonne, IL 60439 Department of Chemistry, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801 United States

† ǁ

Department of Physics and Astronomy, California State University, Northridge, California 91330 United States



Materials Science Division; § Chemical Sciences and Engineering Division; ∞ X-ray Science Division, Advanced Photon Source; Argonne National Laboratory, Argonne, Illinois 60439 United States Keywords: Lithium-sulfur battery, solvate electrolyte, hydrofluoroether cosolvent, sulfur reduction kinetics, in situ Raman spectroscopy ABSTRACT: Li‒S batteries are a promising next-generation battery technology. Due to the formation of soluble polysulfides during cell operation, the electrolyte composition of the cell plays an active role in directing the formation and speciation of the soluble lithium polysulfides. Recently, new classes of electrolytes termed ‘solvates’ that contain stoichiometric quantities of salt and solvent and form a liquid at room temperature have been explored due to their sparingly solvating properties with respect to polysulfides. The viscosity of the solvate electrolytes is understandably high limiting their viability, however, hydrofluoroether cosolvents, thought to be inert to the solvate structure itself, can be introduced to reduce viscosity and enhance diffusion. Nazar and coworkers previously reported that addition of 1,1,2,2-tetrafluoroethyl 2,2,3,3tetrafluoropropyl ether (TTE) to the LiTFSI in acetonitrile solvate, (MeCN)2‒LiTFSI results in enhanced capacity retention compared to the neat solvate. Here, we evaluate the effect of TTE addition on both the electrochemical behavior of the Li-S cell and the solvation structure of the (MeCN)2‒LiTFSI electrolyte. Contrary to previous suggestions, Raman and NMR spectroscopy coupled with ab initio molecular dynamics simulations show that TTE coordinates to Li+ at the expense of MeCN coordination, thereby producing a higher content of free MeCN, a good polysulfide solvent, in the electrolyte. The electrolytes containing a higher free MeCN content facilitate faster polysulfide formation kinetics during the electrochemical reduction of S in a Li-S cell likely as a result of the solvation power of the free MeCN.

INTRODUCTION Investigation of Li-S batteries is motivated by the potential of high gravimetric capacity using inexpensive and earth abundant materials.1 The theoretical capacity of the S cathode is 1675 mAh g⎻1 on a S basis, over 10 times higher than the reversible capacity of the conventional intercalation cathode LiCoO2 (140 mAh g⎻1) used in a Li ion battery.2 One of the challenges of the Li-S battery lies in the complicated redox processes at the S cathode that results in

poor cyclability. Electrochemical reduction of elemental S8 proceeds via the formation of soluble polysulfide species3 whose diffusion from the cathode decreases the Coulombic efficiency and capacity retention of the cell.1 In addition to reduced capacity, crossover of intermediate products causes undesirable side reactions at the Li anode, an effect colloquially known as the polysulfide shuttle.4 Some of the efforts to mitigate polysulfide dissolution include mechanical barriers in the cathode structure to prevent diffusion,5,6 tailored interlayers between the electrodes to prevent crossover,7,8 electrolyte additives to passivate the anode or control the discharge mechanism,9–11 and alternative electrolyte compositions aimed at minimizing polysulfide solubility.12,13 Here, we focus on understanding the effect of minimally solvating electrolytes, with respect to polysulfide solubility, on cell performance. The electrolyte families that exhibit minimal polysulfide solubility do so by utilizing high concentrations of the electrolyte salt to take advantage of the common ion effect. High salt concentration causes the equilibrium between solid Li+ and solvated Li+ to shift toward the solid,14 thereby reducing Li+ solubility. This approach was pioneered by Cho and coworkers who observed a dependence of the cell performance on the concentration of lithium bis(trifluoromethane sulfonyl) imide (LiTFSI) salt, varied from 1-5 M, in a dioxolane (DOL)/dimethoxy ethane (DME)-based electrolyte.14 Increasing the concentration of the LiTFSI higher to 7 moles of salt in 1 L of solvent results in even higher Coulombic efficiencies and better cycling performance.15 In concert with the common ion effect, Watanabe and coworkers have shown that the local solvation structure plays an additional, if not more significant, role in affecting cell performance.16 The local solvation structure of Li+ involves coordination of solvent molecules, thereby reducing the ability of the coordinated solvent to solubilize other solutes.16 When the stoichiometric ratio of salt to solvent is increased to a value in which all solvent molecules are coordinated to Li+ centers in the liquid phase, a “solvate” electrolyte is achieved.17 Although the stoichiometry of the complex suggests that each solvent molecule is part of a Li-complex, an equilibrium between coordinated solvent and free anion vs. free solvent

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

and coordinated anion results in some free solvent molecules in the solution.18 The equilibrium is effected by the TFSI anion, which results in only minimal formation of free solvent in a tetraglyme solvent.18 The disadvantage of the solvate electrolytes lies in the high viscosity and low ionic conductivity relative to conventional salt-in-liquid organic electrolytes. In order to decrease the viscosity of the solvate, Watanabe and coworkers introduced a second solvent that was expected to be inert with respect to the solvate complexes formed by the Li and the gylme solvent molecules.16 Hydrofluoroethers (HFE) are good options to reduce the viscosity of the solvate due to their miscibility with the solvate complexes, inability to solubilize lithium polysulfides, and suggested reductive stability against Li metal.16 There is no evidence in the Raman spectra for HFE coordination directly to the Li+ in the LiTFSI-glyme solvate.19 Further studies have shown, however, that the addition of HFE facilitates the formation of solvent separated ion pairs at the expense of TFSI coordination to the tetraglyme-Li+ complex.19 Cycling a Li-S cell in the HFE/LiTFSI-glyme electrolyte, despite the low solubility of lithium polysulfides in the LiTFSIglyme,20 reveals one high voltage and one intermediate voltage plateau in the discharge profile16 reminiscent of the discharge mechanism observed in traditional lithium polysulfides-solvating electrolytes. To compare, a solid-state conversion to Li2S is expected to result in a single plateau.21 The solvate electrolytes discussed hitherto were prepared with ethereal solvents, a natural extension of the conventional electrolytes for Li-S. Nazar and coworkers, however, evaluated solvate electrolytes for Li-S batteries composed of LiTFSI in acetonitrile (MeCN).22 The stoichiometry of fully complexed, 4-coordinate Li is suggested to be 1 mol LiTFSI and 2 mol MeCN,23 therefore, the solvate electrolyte will be henceforth referred to as (MeCN)2−LiTFSI. Cycling a Li-S cell with the (MeCN)2−LiTFSI electrolyte changes the shape of the discharge and charge profiles relative to the conventional 1 M, ethereal electrolytes.22 The shape of the discharge profile under non-equilibrium conditions is more sloping resulting in less-defined plateaus, however, galvanostatic titration experiments reveal two distinct plateaus on the discharge.22 These changes in the profile suggest that the discharge processes have changed as a result of the electrolyte and indeed in situ X-ray absorption spectroscopy (XANES) shows the formation of solid Li2S earlier in the discharge process relative to the conventional electrolytes.22 MeCN is a counterintuitive solvent for use with Li metal due to its rapid, exothermic decomposition on Li.24 Due to the high salt concentration, however, no decomposition was observed.22 Additionally, around the same time, Yamada and coworkers demonstrated reversible Li+ intercalation into graphite and reversible Li metal electrodeposition and stripping in the (MeCN)2−LiTFSI solvate electrolyte.25 The SEI on the Li metal surface in the (MeCN)2−LiTFSI electrolyte is mostly composed of decomposition products of the anion thereby reducing the kinetics of MeCN decomposition. Additionally, the coordination of MeCN to Li+ inhibits its reactivity with the Li metal.26 Utilizing the (MeCN)2−LiTFSI electrolyte in a Li-S battery results in high capacity and stable cycling relative to the low concentration, ethereal analogues.22 However, this behavior is only attainable when the hydrofluoroether 1,1,2,2,-tetrafluoroethyl 2,2,3,3tetrafluoropropyl ether (TTE) is used to presumably reduce the viscosity of the electrolyte and increase the conductivity.22 The neat solvate electrolyte results in the drastic capacity fade that normally plagues Li-S cells.22 In addition to the TTE’s ability to reduce the electrolyte viscosity, the TTE also affects the SEI on the Li metal surface27–29 and could have an impact on the S reduction pathway.30 Fluorinated ethers have been shown to suppress the polysul-

Page 2 of 13

fide shuttle,31 reduce self-discharge,32 and generally improve the Coulombic efficiency and cycle life.30 In order to understand the beneficial effects of adding TTE to the (MeCN)2−LiTFSI solvate electrolyte on sulfur electrochemistry, we evaluate the electrochemical behavior of the Li-S cell as TTE is added to both the (MeCN)2−LiTFSI solvate electrolyte and a similar electrolyte with slightly lower LiTFSI concentration, (MeCN)2.5−LiTFSI. The nomenclature represents the mole ratios of the components in the solution and not the local solvation structure. The (MeCN)2.5−LiTFSI solution can be visualized as the (MeCN)2−LiTFSI solvate with excess MeCN. A slightly lower LiTFSI concentration is expected to contain higher relative amounts of free MeCN allowing us to decouple effects of viscosity and uncoordinated MeCN content. The S reduction kinetics are evaluated by tracking the Raman modes of dissolved polysulfide as a function of potential using in situ Raman spectroscopy. We observe that addition of TTE to the (MeCN)2.5−LiTFSI electrolyte results in polysulfide formation kinetics similar to those observed in the conventional 1 M, ethereal electrolytes. To understand this effect, the solvation structure of the (MeCN)x−LiTFSI electrolytes was evaluated. Raman and NMR spectroscopy along with ab initio molecular dynamics simulations show that addition of TTE results in some preferential coordination of Li+ to the TTE at the expense of the MeCN, thereby increasing the relative quantity of free MeCN. This finding contradicts the current assumption that the TTE is inert with respect to the Li+ solvation structure. Increased TTE content results in increased free MeCN that is then able to solubilize polysulfides at the cathode likely facilitating faster S electroreduction kinetics. Interestingly, although the kinetics of polysulfide formation is enhanced, the diffusion of the polysulfide into the bulk electrolyte is greatly reduced relative to the conventional electrolytes. EXPERIMENTAL Electrolyte preparation: Lithium bis(trifluoromethylsulfonyl imide) (LiTFSI) salt was purchased from Sigma Aldrich and dried at 130°C under vacuum for 8 hrs. Nominally anhydrous acetonitrile (MeCN, 99.8%, Sigma Aldrich) and 1,1,2,2-tetrafluoroethyl 2,2,3,3-tetrafluoropropyl ether (TTE, 99%, Synquest Laboratories) were further dried over activated alumina for at least 7 days. The solvate electrolyte, (MeCN)2−LiTFSI, was prepared from MeCN and LiTFSI in a mol ratio of 2:1 followed by stirring overnight to yield a clear, colorless solution with no residual solid. The density of the (MeCN)2−LiTFSI was measured in a 1.00 mL volumetric − flask as 1.47 g mL 1 corresponding to a LiTFSI concentration of 3.98 M. The (MeCN)2.5−LiTFSI solution was prepared similarly and resulted in a concentration of 3.70 M. The LiTFSI concentration series was prepared by diluting a stock solution of 3.70 M LiTFSI in MeCN. The 3.70 M LiTFSI in MeCN was prepared in a 10 mL volumetric flask in an Ar glove box by weighing in 10.6223 g LiTFSI and then diluting with MeCN. The electrolytes with TTE added were prepared by diluting (MeCN)2−LiTFSI electrolyte (3.98 M) or (MeCN)2.5−LiTFSI electrolyte (3.70 M) with TTE to obtain (MeCN)x−LiTFSI:TTE at 2:1, 1:1, and 1:2 volume ratios. The water content of each electrolyte measured by Karl Fisher titration (Photovolt Aquatest Karl-Fischer Coulometric Titrator) is less than 15 ppm. The stability of the electrolytes vs. the Li metal anode was evaluated by soaking Li metal in the electrolytes and observing color changes. The electrolytes turn from colorless to yellow when electrolyte decomposition occurs25 and no color change was observed for any of the dried electrolytes tested. This experiment was done previously by Yamada et al. to show the compatibility of the electrolytes with Li metal.25 We note, however, that significant

ACS Paragon Plus Environment

Page 3 of 13

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

ACS Applied Materials & Interfaces

electrolyte decomposition occurs in electrolytes prepared with relatively wet, as−purchased LiTFSI. Without drying the salt, the solvate electrolytes contain >50 ppm water and show significant yellowing when in contact with Li metal. 7Li and 19F NMR confirming electrolyte decomposition are shown in the SI. Two-electrode Swagelok cell preparation: The cathode material was prepared by melt diffusion of S (99.98%, Sigma-Aldrich) in ordered mesoporous carbon, CMK‒3 (BET1000, ACS MATERIAL), to achieve 50% S@CMK‒3 as described previously.5,22 The cathode slurry consisting of 80 wt% S@CMK‒3 nanocomposite, 10 wt% carbon black (Super P Li, Timcal Inc.), and 10 wt% polyvinylidene fluoride (PVDF, Kynar 2801) binder was mixed with anhydrous N-methyl-2-pyrrolidone (NMP, SigmaAldrich). S loading in the resulting cast was 0.28−0.4 mg cm−2. Li‒ S cells were assembled in a modified Swagelok tube apparatus (nylon, 0.5” inner diameter, Chicago Fluid System Technologies). The Swagelok cell consisted of a Li metal anode (99.99%, Alfa Aesar), a Whatman glass fiber separator, and the S@CMK‒3 cathode. Electrolyte volume in all cells was controlled at about 0.03 mL.33 The ratio of sulfur to electrolyte volume is ~13 g/L. Characterization details Cyclic voltammetry (CV): CV experiments were performed in a Swagelok cell scanning from 3.2 V to 1.5 V (vs. Li/Li+) at 0.1 mV s⎻1 with a CH Instruments potentiostat. The cells were allowed to rest at open circuit for at least 12 hr before cycling to allow for electrode wetting. All potentials are referenced to Li/Li+. Raman spectroscopy measurements: Raman spectroscopy was performed using a system described previously.10,34 The typical acquisition time was 150s per spectrum for in situ Raman spectroscopy and 240s per spectrum for ex situ Raman spectroscopy. Excitation was provided by a 20 mW 632 nm He-Ne laser. The instrumental resolution was ca. 6 cm⎻1 over the spectral range of the measurement. 7 Li NMR: All 7Li NMR spectra were measured on a 600 MHz Varian instrument in 5 mm screw cap NMR tubes at the University of Illinois at Urbana-Champaign School of Chemical Sciences NMR facility. An external standard in the form of a sealed, coaxial capillary containing 10 M LiCl in D2O was introduced in all samples. The resonance of the 10 M LiCl in D2O was referenced to 1 M LiCl in D2O. All reported 7Li resonances are referenced to 1 M LiCl. The 90° pulse was measured for each sample and at every temperature. The 7Li longitudinal relaxation time, T1, was measured with a standard inversion-recovery pulse sequence with varying τ values, the time delay between the 90° and 180° pulses. The T1 was then calculated by determining the time constant relating the strength of the FID as a function of τ using exponential curve fitting. The 7Li spectra were then measured using the calibrated 90° pulse width, a relaxation delay equal to 5 times the longest T1, and an acquisition time of 5 s. Ab Initio Molecular Dynamics (AIMD) simulation: The bulk electrolytes were simulated using a density functional theory (DFT) formalism. For all bulk electrolyte simulations, a simulation cell (18 Å x 18 Å x 18 Å) consisting of about 400 atoms was used to represent the electrolyte. Four different systems were examined including a dilute LiTFSI/MeCN electrolyte (1.7 M LiTFSI in MeCN), a concentrated (MeCN)2−LiTFSI solvate electrolyte (3.98 M LiTFSI in MeCN), and a (MeCN)2−LiTFSI:TTE system at volume ratios of 1:1, and 1:2. Concentrations were changed by varying the TFSI:MeCN ratio. In all simulations, the electrolyte density was fixed close to the experimentally measured values. All the AIMD calculations were carried out using the CPMD program package35 using periodic boundary conditions, planewave

basis sets as the expansion of electronic wavefunctions, and atomic pseudopotentials. In particular, one-electron orbitals were expanded in a planewave basis with a kinetic energy cutoff of 45 Ry restricted to the Gamma point of the Brillouin zone. Medium soft norm-conserving Trouiller-Martins pseudo-potentials were selected for all the elements in the generalized gradient approximation.36 All AIMD simulations utilized the PBE functional within the spinpolarized Kohn-Sham formalism with the Grimme dispersion correction for van der Waals interactions.37 The energy expectation values were computed in reciprocal space using the KleinmanBylander transformation.38 Starting from DFT optimized geometries, the system was thermally equilibrated at T = 300 K using the Nose-Hoover thermostat via canonical ensemble (NVT)39 prior to the production run. Throughout the simulation, a molecular dynamics time step of 3 atomic units was used and all production runs of AIMD trajectories were obtained after ~ 1 – 1.2 ps of thermal equilibration. Quantum Chemistry simulation: The thermally equilibrated structures obtained from the AIMD simulation were further interrogated for their electronic, thermochemical and NMR properties by using DFT calculations as implemented in the Gaussian 09 code40 with the B3LYP hybrid functional.41 The thermally equilibrated structures were geometry optimized by using the 6-31G(d,p) basis set. The structures chosen for the geometry optimization are shown in the SI (Figure S4). NMR properties of these optimized geometries were calculated using the 6-311+G(d,p) basis set. For the NMR calculations, the effects of electrolyte is modeled by using the polarizable continuum model (PCM) at the B3LYP/6-311+G(d,p) level. A dielectric constant (ε) of 35.69 was used in all the PCM calculations assuming the solvent is MeCN. RESULTS AND DISCUSSION Electrochemical behavior of Li-S cells with the MeCN solvate electrolyte We first discuss the effect of electrolyte composition on the electrochemical behavior of the Li‒S cell. Figure 1a shows the CV (cycle 1) of S infiltrated mesoporous carbon (S@CMK‒3) cathodes obtained in a Swagelok cell at a scan rate of 0.1 mV s‒1 cycled from 3.2 V to

Figure 1. CV (cycle 1) of Li‒S cells prepared with a Li metal anode, S@CMK‒3 cathode slurry, and the indicated electrolyte cycled at 0.1 mV s‒1. (a) (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI are shown

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

with the CV of the conventional electrolyte, 1 M LiTFSI in DOL/TEGDME, overlaid for comparison. (b) The first CV cycle of

Page 4 of 13

(MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI diluted with TTE at a 1:1 volume ratio.

Figure 2. In situ Raman spectra of the as-prepare sulfur-carbon cathode shown at representative potentials with (a) 1 M LiTFSI in DOL/TEGDME, (b) (MeCN)2–LiTFSI, (c) (MeCN)2–LiTFSI:TTE (1:1), and (d) (MeCN)2.5–LiTFSI:TTE (1:1). The Raman mode associated with S3•− is highlighted. (e) The potential profile and (f) the normalized Raman intensity of S3•− were plotted as a function of time obtained in various electrolyte systems. 1.5 V. The electrolytes in Figure 1a include the conventional electrolyte, 1 M LiTFSI in DOL/TEGDME, and two concentrations of LiTFSI in MeCN including (MeCN)2.5−LiTFSI (3.70 M LiTFSI) and (MeCN)2−LiTFSI (3.98 M LiTFSI). The conventional electrolyte CV is shown to facilitate comparison between it and the LiTFSI in MeCN electrolytes. The CV of the LiTFSI in MeCN electrolyte is much different than that obtained in the 1 M LiTFSI in DOL/TEGDME electrolyte. For example, the reduction and oxidation peaks are shifted to more negative and positive potentials, respectively. The potential of the reduction peaks in the MeCN electrolytes is shifted to ~2.15 V and ~1.7 V while the two reduction peaks found in DOL/ TEGDME electrolyte are at 2.4 V and 2.05 V.10,42 In the anodic scan, the two oxidation peaks in the MeCN electrolytes are observed at ~2.45 V and ~2.6 V while the two oxidation peaks in the DOL/TEGDME electrolyte are at 2.4 V and 2.5 V. Such changes could be due to either kinetic or thermodynamic considerations. Previous studies suggest that the kinetics of the S reduction are hindered by the poor solubility of intermediate polysulfides, although the formation of polysulfide intermediates is still observed.22 The cause of poor kinetics is suggested to arise from the inability of MeCN to act as a polysulfide solvent due to its coordination to Li.22 However, many other variables can have an effect on the CV. For example, the activity of Li is different from the MeCN electrolytes to the conventional electrolyte due to the drastic change in concentration (1 M to nearly 4 M) and the viscosity of the MeCN electrolytes is much greater.22 The CV was also measured for the (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI electrolytes with addition of the TTE cosolvent. Figure 1b shows the first CV cycle of (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI diluted with TTE at a 1:1 volume ratio. The potentials of the reduction peaks are at ~2.05 and ~1.85 V in both cases. Interestingly, addition of TTE causes the first reduction peak to shift to more negative potentials in both of the MeCN electrolytes. This result is surprising because addition of TTE decreases

the viscosity of the solvent from 138 cP in (MeCN)2−LiTFSI to 8.6 cP in (MeCN)2−LiTFSI:TTE = 1:122 which we would expect to enhance the kinetics of S reduction resulting in a more positive reduction potential. As a reference, the viscosity of the conventional electrolyte, 1 M LiTFSI in DOL/TEGDME, is ~3cP.43 Such an effect is observed for the second reduction peak which is shifted to a more positive potential as a result of TTE addition. Further, the current density of the redox peaks decreases with cycling (Figure S2), which suggests either active material loss or the presence of a decomposition event. Investigation of S reduction kinetics in the solvate electrolyte with TTE To understand the formation kinetics of the polysulfide species and the effect of LiTFSI concentration on the stability of polysulfides at the electrode surface, we used in situ Raman spectroscopy. The stability of polysulfide species was measured by tracking the intensity of the respective Raman modes as a function of time after applying a reducing bias. Figure 2 shows in situ Raman spectra of the asprepare sulfur-carbon cathode shown at representative potentials with (a) 1 M LiTFSI in DOL/TEGDME, (b) (MeCN)2–LiTFSI, (c) (MeCN)2–LiTFSI:TTE (1:1), and (d) (MeCN)2.5–LiTFSI:TTE (1:1). The potential profile for each cell is shown in Figure 2e and the Raman intensity of S3•− as a function of time in the various electrolyte systems is shown in Figure 2f. The formation of S3•− after a reducing bias is applied is a result of the electrochemical reduction of S, but could also be due to disproportionation reactions such as S62− 2 S3•−.42,44–46 Figure 2a shows in-situ Raman spectra of the sulfur cathode obtained at potentials between 2.8 V and 2.2 V in the electrolyte with DOL/TEGDME. As expected, the Raman band a associated with radical anion S3•− is observed at 2.2 V.10,47 In the solvate electrolyte, Figure 2b-d shows in-situ Raman spectra of the sulfur cathode

ACS Paragon Plus Environment

Page 5 of 13

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

ACS Applied Materials & Interfaces LiTFSI:TTE electrolyte.

obtained at potentials between 2.8 V and 1.8 V. The short chain polysulfides such as S3•− are formed at 1.8 V. Indeed, the formation of short chain polysulfides in acetonitrile has been suggested by UV-Vis absorption spectroscopy.48 The intensity of band a obtained in the electrolyte with (MeCN)2–LiTFSI and (MeCN)2–LiTFSI:TTE (1:1) is lower than the intensity of band a obtained in the electrolyte with (MeCN)2.5–LiTFSI:TTE (1:1), suggesting that more polysulfides are formed near the cathode surface in (MeCN)2.5– LiTFSI:TTE (1:1).

Understanding the solvation structure of Li+ In an attempt to understand the origin of the S3•− persistence in the (MeCN)2−LiTFSI:TTE electrolyte and the enhanced polysulfide formation kinetics in the (MeCN)2.5–LiTFSI:TTE (1:1) system, we evaluated the local structure of the solvents and salt in the electrolyte solution. The local structure can affect the electrochemistry of the Li-S cell by influencing the formation of solvated intermediates. We

We next hold the potential at partial discharge to study the persistence of the different polysulfide species. The potential was stepped from the 2.8 V to the voltage of the second reduction peak to observe the formation of reduction products. Figure 2e-f shows the potential profile and normalized intensity of the Raman bands associated with short chain polysulfides such as S3•− plotted as a function of time after stepping the potential from 2.8 V to 2.2 V in the electrolyte with DOL/TEGDME and 1.8 V in the electrolyte with (MeCN)2-LiTFSI, (MeCN)2-LiTFSI:TTE (1:1) and (MeCN)2.5LiTFSI:TTE (1:1). In the DOL/TEGDME electrolyte, the peak intensity associated S3•− reaches a maximum ca. 20 min after the potential was switched to 2.2 V and the peak intensity stays constant for the next 20 min of the measurement. This behavior suggests that the polysulfide formed at this potential is stable under potential control, at least over this time scale. The peak intensity then decreases by 50% in the following 20 min. A decrease in the peak intensity could be due to either diffusion of the S3•− away from the cathode surface or S3•− reduction. However, when the cathode is held at 2.2 V for an extended time (ca. 2 hr), the intensity of the S3•− Raman mode remains at ~15% of the maximum value (data not shown).

Figure 3. The Raman spectra of (a) (MeCN)2.5−LiTFSI and (b) (MeCN)2−LiTFSI. The spectra of the (MeCN)x−LiTFSI diluted with hydrofluoroether TTE at volume ratios of 1:2, 1:1, and 2:1 are also shown. The lower wavenumber region shows the free TFSI mode, mode a, and coordinated TFSI, mode b. The higher wavenumber reveals two MeCN peaks. The higher wavenumber region shows MeCN modes c, d, e, and f. Modes d and f arise due to shifts in modes c and e, respectively, due to coordination to Li+. (c) As TTE is titrated into both the (MeCN)2.5−LiTFSI and the (MeCN)2−LiTFSI solution, more free MeCN is observed as determined by the ratio of the peak areas of the free MeCN mode, mode c, to coordinated MeCN, mode d.

The intensity of S3•−-associated peak formed in the (MeCN)2−LiTFSI electrolyte and the (MeCN)2−LiTFSI:TTE (1:1) electrolyte reaches a maximum ca. 45 min after the potential was switched to 1.8 V. This behavior shows that the formation rate of polysulfides in the solvate electrolyte is slower than the formation rate of polysulfides in the DOL/TEGDME. The peak intensity then drops slightly to 80% of the maximum value and remains at 80% over a 2 hr period under potential control. The persistence of the Raman peak intensity associated with the short chain polysulfide suggests that the consumption and diffusion of polysulfide at 1.8 V (the second reduction peak) is not substantial. Thus, the S3•− is more stable at the cathode surface in the solvate electrolytes, with or without TTE, compared to the conventional DOL/TEGDME electrolyte.

discuss the solvation structure as probed by Raman and NMR spectroscopy coupled with ab intio molecular dynamics simulations.

We also examined the (MeCN)2.5–LiTFSI:TTE (1:1) electrolyte. Figure 2e shows that the persistence of S3•− is similar to other electrolytes containing LiTFSI and MeCN. Interestingly, however, the appearance of the S3•−-related bands occurs as quickly as it does in DOL/TEGDME, which suggests that polysulfide formation kinetics are facilitated in this electrolyte relative to the (MeCN)2–

Raman spectroscopy Coordination to Li+ centers can cause measurable shifts in the Raman modes associated with the solvent or anions. The Raman spectra of (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI both neat and with additions of TTE at volume ratios of 2:1, 1:1, and 1:2 are

Table 1. Assignments of Raman modes in the LiTFSI concentration series and the (MeCN)x-LiTFSI electrolytes with TTE diluents. Raman shift (cm−1) ID

(MeCN)2.5−LiTFSI:TTE vol. ratio 0

2:1

1:1

1:2

Assignment

(MeCN)2−LiTFSI:TTE vol. ratio 0

2:1

1:1

Species

Mode(s)

Ref.

1:2

a

739

739

739

739

739

739

739

739

TFSI

S-N stretch + C-S stretch + CF3 bend

49–52

b

746

746

746

746

746

746

746

746

TFSI

mode a, coordinated to Li+

49–53

c

2255

2256

2256

2257

2253

2254

2255

2255

MeCN

C≡N stretch

54,55 +

d

2278

2279

2280

2280

2278

2278

2279

2280

MeCN

mode c, coordinated to Li

e

2294

2295

2296

2296

2296

2297

2297

2297

MeCN

C−H stretch

f

2308

2309

2309

2309

2308

2309

2309

2309

MeCN

mode e, coordinated to Li+

ACS Paragon Plus Environment

23,53,56,57 54,55

ACS Applied Materials & Interfaces

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

shown in Figure 3. Assignments for the Raman modes in Figure 3 are given in Table 1. Figure 3a and 3b shows the TFSI, modes a and b, and MeCN modes, modes c-f, that are most sensitive to Li+ coordination. Both TFSI and MeCN are coordinating to Li+ in these solvate electrolytes and the anion modes will be discussed first. The TFSI mode a is a combination of S−N stretching, C−S stretching, and CF3 bending.49–52 TFSI expresses many additional modes, however, the other TFSI modes are either convolved with TTE bands or affected by the distribution of conformers in solution.51 Mode a, however, is sensitive to ionic interactions making it a good probe for coordination.51 Upon coordination of the TFSI to Li+ through the sulfonyl O, mode a shifts to higher wavenumbers resulting in a new mode,49–51 labeled mode b in Figure 3a and 3b. At low concentrations of LiTFSI in MeCN, the TFSI Raman peak around 740 cm−1 is dominated by contributions from uncoordinated TFSI due to the formation of solvent separated ion pairs.53 As the concentration of LiTFSI is increased, the new peak associated with coordinated TFSI, mode b, grows in at the expense of mode a.53 We observe contributions from coordinated TFSI above 2 M LiTFSI (see SI). The TFSI mode for both (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI is largely dominated by mode b, coordinated TFSI. Although TFSI is generally considered to be a weakly coordinating anion, the high concentration favors the formation of contact ion pairs over the solvent separated ions pairs (SSIP) observed at lower concentrations.53 As TTE is added to both (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI solutions the TFSI mode remains largely unchanged indicating that TFSI coordination is unaffected by the TTE solvent, even though the TFSI concentration has decreased. The spectra are overlaid and normalized in Figure 3a and 3b to facilitate comparison of the neat (MeCN)x−LiTFSI solutions with the addition of TTE. MeCN coordination was interrogated by evaluating the C≡N stretching mode at 2252 cm−1, labeled mode c in Figure 3a and 3b. The MeCN solvent coordinates to Li+ through partial donation of the lone pair on the N resulting in a shift of the C≡N stretch to higher wavenumbers.23,56,57 The coordinated C≡N stretch is labeled mode d. The C−H stretching mode, mode e, is also shown in Figure 3a and 3b. A new mode, mode f, is observed at high LiTFSI concentrations likely due to perturbations in the C-H stretch caused by Li+ coordination. The C−H modes, however, are understandably less diagnostic of Li+ coordination and will not be discussed further. The quantity of free MeCN in the electrolyte is important when considering the functionality of the solvate electrolyte vs. a Li metal anode as MeCN readily and exothermically decomposes on Li metal.58 Free MeCN, however, is not necessarily decomposed on the Li surface in the solvate electrolyte due to the formation of an SEI that prevents or limits MeCN decomposition.26 Due to changes in the electronic structure of the TFSI anion as a result of (MeCN)2−LiTFSI complexes, the TFSI is preferentially decomposed on the Li surface resulting in an SEI that is mostly composed of inorganic, anion decomposition products.26 The SEI is believed to hinder the decomposition of any free MeCN. Although MeCN may not decompose at the anode, consideration must be given to the effect of free solvent molecules on the reaction pathway of S at the cathode. The sensitivity of the C≡N stretching mode to Li+ coordination allows for the determination of relative changes in the population of MeCN molecules coordinated to Li+ relative to those that are behaving as free, or bulk, MeCN. Figure 3a and 3b shows the MeCN modes of the (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI solutions neat and diluted with TTE at volume ratios of 2:1, 1:1, and 1:2. The spectra are overlaid and normalized to mode d. The

Page 6 of 13

two neat electrolytes, (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI, contain both free MeCN, mode c, and coordinated MeCN, mode d. The presence of free MeCN in the (MeCN)2−LiTFSI solution suggests that in order to satisfy a 4 coordinate Li cation, some TFSI ligands must be coordinated to more than one Li+ forming aggregate structures that have been proposed previously.53 The formation of aggregates should be manifested in the Raman modes of the TFSI. However, the wavenumber shift of TFSI coordinated to one Li+, mode b, relative to that of TFSI coordinated to two Li+ is expected to be very small and not resolved.53 As the TTE content is increased in both electrolytes, an increase in the intensity of mode c relative to mode d is observed suggesting that TTE addition increases the amount of free MeCN. The effect of TTE addition on MeCN coordination suggests that the

Figure 4. (a) 7Li NMR of the LiTFSI concentration series in MeCN at 3.98 M, 3.70 M, 3.08 M, 2.47 M, 1.85 M, 0.93 M, and 0.46 M. The (MeCN)2−LiTFSI and (MeCN)2.5−LiTFSI electrolytes are at LiTFSI concentrations of 3.98 M and 3.7 M, respectively. The (b) 7 Li chemical shift, (c) FWHM, and (d) T1 of each solution plotted as a function of LiTFSI concentration. TTE is active with respect to the solvate structure by displacing coordinated MeCN from the first solvation shell. The Raman data alone, however, is unable to determine if the TTE itself is involved in the Li+ first solvation shell. The relative quantity of free MeCN vs. coordinated MeCN is related to the ratio of the Raman peak areas of mode c vs. mode d. To compare the relative quantity of free vs. coordinated MeCN, the Raman spectra were fit with component Gaussian functions (shown in the SI) and the peak area ratio of free MeCN vs. coordinated MeCN was calculated. The peak area ratios are shown in Figure 3c for both (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI as a function of TTE content. As expected, the (MeCN)2.5−LiTFSI solution contains a higher relative quantity of free MeCN vs. coordinated MeCN compared to the (MeCN)2−LiTFSI solution due to the low concentration of LiTFSI. Interestingly, as TTE is titrated into the (MeCN)2.5−LiTFSI electrolyte relatively more MeCN is released from the solvate structure compared to the (MeCN)2−LiTFSI electrolyte as evidenced by the higher slope show in Figure 3c. 7

Li NMR Spectroscopy

ACS Paragon Plus Environment

Page 7 of 13

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

ACS Applied Materials & Interfaces

7 Li NMR was measured for the solvate electrolytes to gain insight into the solvation structure of the Li. The 7Li NMR of a series of LiTFSI concentrations in MeCN up to the (MeCN)2.5−LiTFSI and (MeCN)2−LiTFSI compositions at [LiTFSI] = 3.7 M and [LiTFSI] = 3.98 M, respectively, will be discussed first. Figure 4a shows the 7Li NMR spectra and Figure 4b and 4c shows the 7Li chemical shift and full width at half maximum (FWHM) as a function of LiTFSI concentration. The chemical shift reveals changes in the chemical environment of the Li nucleus as a result of influences by the first solvation shell causing either deshielding, a shift to more positive ppm values, or a shielding, a shift to more negative ppm values, of the Li nucleus.59 Li complexes with higher coordination numbers, increasingly defined local structure, and more electronegative ligands will result in a deshielded environment. 7Li is a quadrupolar nucleus with s = 3/2 which causes the peak width to change due to changes in the symmetry of the electron density around the Li nucleus. More

Figure 5. 7Li NMR of the (a) (MeCN)2.5−LiTFSI and (b) (MeCN)2−LiTFSI electrolyte with addition of TTE at volume ratios of 2:1, 1:1, and 2:1 (MeCN)x−LiTFSI:TTE. The (c) 7Li chemical shift, (d) full width at half maximum, and (e) T1 of each solution plotted as a function of TTE content. symmetric environments will yield narrower line widths. Therefore, both the chemical shift and the peak width yield information regarding the solvation structure. In the dilute regime, up to about [LiTFSI] = 2 M, the chemical shift and peak width are relatively independent of LiTFSI concentration. The negligible change in chemical shift suggests that the first solvation shell remains unchanged with concentration. In the dilute regime, we can assume the solution contains solventseparated ion pairs as demonstrated by Henderson and cowork53 ers. The concentration of LiTFSI does not affect the 7Li resonance because the solvent separated Li is maintained. The negligible change in 7Li chemical shift in the low TFSI concentration region is counter to results observed for other anions at concentrations below 0.5 M including bromide, iodide, perchlorate, and tetraphenyl borate in MeCN59 confirming that TFSI is a relatively weaker coordinating ligand. Additionally, the FWHM remains relatively constant and narrow due to the fast tumbling of solvent molecules around the Li center resulting in a symmetric, radial contribution to the local bonding environment. The formation of

SSIP at low concentrations is in good agreement with prior Raman results.53 Above [LiTFSI] = 2 M, the chemical shift and the FWHM change as the concentration of LiTFSI increases suggesting that the local coordination around the Li is changing. Therefore, the solution is no longer characterized by SSIPs and instead begins to form contact ion pairs in which the TFSI is coordinated to the Li centers. Due to the multiplicity of possible structures in solution, the chemical shift changes consistently with concentration until the fully saturated (MeCN)2−LiTFSI solution is achieved. Concurrently, the FWHM broadens due to the formation of locally defined complexes that constrain the position of the ligands in the first solvation sphere (Figure 4c). The loss of rotational and translational freedom of the molecules in the first solvation sphere causes anisotropy in the Li bonding environment which is manifested by a broader 7Li signal. Additionally, larger LiTFSI concentrations will result in more extensive dipolar interactions due to the formation of contact ion pairs which will also contribute to the broadening of the resonance.

Figure 6. Variable temperature 7Li T1 curves for the solvate, (MeCN)2−LiTFSI, and the solvate diluted with TTE. The longitudinal relaxation time, T1, was measured as a function of LiTFSI concentration at room temperature (23 °C) and is shown in Figure 4d. The T1 reports on decay constant associated with the nuclear spin relaxing to its equilibrium position in a magnetic field. The nucleus relaxes by exchanging energy with its surroundings via translational and rotational motions.60 Quadrupolar nuclei, however, are able to distribute energy through quadrupolar relaxation mechanisms and these mechanisms are assumed to be 62 the dominate mechanism.60,61, Thus, for a quadrupolar nucleus, the T1 reports on the symmetry of the local bonding environment around Li. As the LiTFSI concentration is increased in the MeCN solutions, the T1 decreases until the highest concentration and the lowest T1 are achieved. At low concentrations, the fast tumbling of solvent molecules around the solvent separated Li+ results in a symmetric bonding environment leading to a relatively slow T1. The drop in T1 as a function of concentration suggests that the Li+ is continually forming more well-defined complexes forcing asymmetric geometries around Li. As the LiTFSI concentration is increased, the viscosity of the solution increases significantly, which can also contribute to the T1. Similar relationships have been ob-

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

served for the 7Li T1 with increasing concentration of LiBF4 in propylene carbonate.63 The decrease in T1 correlates well with the understanding that as the solution LiTFSI concentration is increased, contact ion pairs are formed. 7

Li NMR was measured for the (MeCN)2−LiTFSI and (MeCN)2.5−LiTFSI solutions with addition of TTE and is shown in Figure 5. The addition of TTE to both the (MeCN)2−LiTFSI and (MeCN)2.5−LiTFSI solutions causes a shift of the 7Li resonance to a more deshielded region (Figure 5a and 5b), a trend opposite to that observed when MeCN is used to dilute the solvate. Interestingly, although the viscosity of the solution is decreased as TTE is added, the peak width remains broader compared to the same concentrations with only MeCN. Thus, the asymmetry caused by the defined solvation structure around the Li center is largely maintained as TTE is added. Slight linewidth narrowing is observed as TTE is added, however, the magnitude of the FWHM change is relatively small. Additionally, the T1 remains short after TTE additions at all volume ratios suggesting minimal changes on asymmetry of the local coordination of the Li+ despite its nominally lower concentration of LiTFSI. Although the T1 and the FWHM do not exhibit significant changes as TTE is added to the two electrolyte solutions, the 7Li chemical shift does shift to more deshielded resonances as TTE is added. From the Raman spectra, we know that TTE addition causes an increase in free MeCN and hence a decrease in MeCN coordination, however, lower coordination of the Li would likely result in a shift of the 7Li resonance to a more shielded region and an increase in T1. Therefore, the shift of the 7Li resonance could be due to the inclusion of a more electronegative ligand relative to the solvation structure with no TTE, such as the F in TTE. To confirm changes in the local structure, variable temperature T1 curves were measured for the neat (MeCN)2−LiTFSI solution and (MeCN)2−LiTFSI diluted with TTE (Figure 6). The 7Li T1 reports on both a change in local environment around the Li+ and long range interactions such as diffusional and rotational motion.60,64 Variable temperature T1 experiments can decouple these two effects. A shift in the location of the minimum T1 to shorter times indicates a change in local coordination while a shift of the minimum to a lower temperature is due to increased mobility.60 The T1 minimum is observed in the temperature range studied for the (MeCN)2−LiTFSI solution at 40 °C of 0.55 s. Upon addition of TTE, the minimum T1 shifts to lower temperatures indicating a change in mobility presumably due to the lower viscosity. Concurrently, however, addition of TTE also lowers the minimum T1 suggesting a change in coordination of the Li+. This effect confirms the previous assumption that the deshielding of the Li is due to a change in the coordination of Li, shown in Figure 5a and 5b. The coordination of the solvate complex formed in the (MeCN)2−LiTFSI solution remains largely intact after TTE addition, as the Raman also suggests. However, a shift in the 7Li resonance to a more deshielded region in conjunction with a change in the T1 minimum to faster time scales suggests that the local structure is changing as TTE is added to the solvate. The deshieldeding suggests that one of the ligands involved in the (MeCN)2−LiTFSI complex, either the MeCN or the TFSI, is being replaced by the more electronegative TTE. Raman spectroscopy suggests that the inclusion of TTE in the first solvation shell comes at the expense of the MeCN coordination, leading to more free MeCN in solution as the TTE concentration is increased. Ab Initio Molecular Dynamics Simulations

Page 8 of 13

In order to further examine Li coordination to TFSI and MeCN along with putative TTE involvement in the Li coordination environment, we performed ab initio molecular dynamics (AIMD) simulations on the (MeCN)2‒LiTFSI complex and quantum chemical calculations as described in the Experimental Section. We evaluated the local coordination of Li+ in the first solvation shell by calculating the integrated radial distribution function from thermally equilibrated AIMD trajectories and measuring the coordination number of each correlation at a distance of 2.7 Å from Li+. Figure 7 shows the average local coordination around Li+ in 1.7 M LiTFSI in MeCN, (MeCN)2‒LiTFSI (3.98 M LiTFSI in MeCN), and (MeCN)2‒LiTFSI diluted with TTE at volume ratios of 1:1 and 1:2. The radial distribution function of the 1.7 M LiTFSI in MeCN solution suggests that the Li+ first solvation shell is dominated by Li‒O and Li‒N interactions. The Li‒N coordination results from MeCN ligation while the Li‒O interaction is due to coordination through the sulfonyl in the TFSI anion. The first solvation shell of Li is dominated by nLi-O ~1.7 and nLi-N ~2.3 at 2.7 Å suggesting the formation of a mixed (MeCN)3‒Li+‒TFSI- and (MeCN)2‒LiTFSIlike complex. The 1.7 M LiTFSI solution generally consists of tetrahedrally coordinated Li+.

Figure 7. The computed Li+‒coordination number of (a) 1.7 M LiTFSI in MeCN, (b) (MeCN)2‒LiTFSI (3.98 M LiTFSI in MeCN), and (MeCN)2‒LiTFSI diluted with TTE at (c) 1:1 and (d) 1:2 volume ratios. The first solvation shell in all cases is defined as within 2.7 Å. The correlations in the Li+ first solvation shell change as the concentration of LiTFSI is increased. The integrated radial distribution function of the high LiTFSI concentration (MeCN)2‒ LiTFSI solution is shown in Figure 7b. The total coordination number of Li remains 4, however, the Li‒O coordination number increases from nLi-O ~1.7 in the low concentration solution to nLi-O ~2.1 in the (MeCN)2‒LiTFSI solution suggesting a higher degree of TFSI association. This increased TFSI interaction is in good agreement with previous Raman studies.53 The Li‒N coordination number decreases from nLi-N ~2.3 in the 1.7 M LiTFSI solution to nLi-N ~1.85 in the (MeCN)2‒LiTFSI solution. The coordination numbers obtained in the (MeCN)2‒LiTFSI solution are closer to those expected for the stoichiometric complex, i.e. nLi-O = 2 and nLi-N = 2, suggesting that a majority of complexes in solution resemble the stoichiometric complex. The Li‒N coordination number is

ACS Paragon Plus Environment

Page 9 of 13

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

ACS Applied Materials & Interfaces

slightly less than 2 suggesting undercoordination of MeCN. Therefore, some MeCN must be uncoordinated and behaving as free MeCN. This result agrees well with the Raman data, which shows the free solvent modes even in the absence of TTE. The undercoordination of MeCN is compensated by a slightly overcoordinated Li‒O to maintain a 4 coordinate average for Li+. Coordination numbers >2 can be achieved through the formation of aggregate structures in which one TFSI binds to more than one Li+ center. Aggregate structures have been proposed previously.53 The radial distribution function of the (MeCN)2‒LiTFSI solvate with TTE addition at volume ratios of 1:1 and 1:2 are shown in Figure 7c and 7d, respectively. Upon TTE addition, the coordination number associated with Li‒O initially measured for (MeCN)2‒ LiTFSI remains constant at ~2 at both TTE concentrations. The inability of TTE addition to effect TFSI coordination is confirmed experimentally by the negligible change in the TFSI Raman modes as TTE is added to the (MeCN)2‒LiTFSI (Figure 5). The coordination number of Li‒N, however, decreases significantly as TTE is added to the (MeCN)2‒LiTFSI system. In the neat (MeCN)2‒ LiTFSI solvate, nLi-N ~1.85 and addition of TTE decreases nLi-N to ~1.33 in the 1:1 solution and ~0.82 in the 1:2 solution. The drop in MeCN coordination as a result of TTE addition is in good agreement with the experimental Raman result showing higher free MeCN content as TTE is added to the solvate solution. The decrease in MeCN coordination is compensated by the formation of a new correlation in the first solvation shell associated with Li‒F. The Li‒F interaction that penetrates into Li+ first solvation shell is found to be nLi-F ~0.48 in the 1:1 solution and increases to nLi-F ~1.16 as the TTE content is increased in the 1:2 solution. The Li‒F correlation is due to coordination of the Li+ with the TTE solvent. Formation of TTE coordinated complexes is reasonable as the DFT computed solvation free energy of a MeCN‒LiTFSI‒TTE complex is ΔGsolv ~ ‒0.30 eV which is comparable to that of (MeCN)2‒ LiTFSI, ΔGsolv ~ ‒0.25 eV. Therefore, a competition between TTE and MeCN for Li+ coordination is expected as the concentration of TTE is varied in solution. The 7Li NMR chemical shifts of the complexes were calculated by DFT using the clusters derived from the AIMD simulations (see SI). Increasing the LiTFSI to MeCN ratio results in deshielding of the 7Li resonance consistent with the experimental observation. Furthermore, the DFT calculated 7Li chemical shift of the MeCN‒ LiTFSI‒TTE complex is deshielded relative to the neat solvate even though the coordination number is 4 in both complexes. Therefore, the addition of Li‒F coordination in the first solvation shell of Li+ at the expense of Li‒N causes a deshielding of the Li+ nucleus. The 7Li resonance in the MeCN‒LiTFSI‒TTE complex is shifted +0.30 ppm relative to the 7Li chemical shift of the (MeCN)2‒ LiTFSI complex. The experimentally measured 7Li chemical shifts as TTE is added the solvate electrolyte (~0.1 ppm) are smaller than the predicted value of the stoichiometric MeCN‒LiTFSI‒TTE complex. Experimentally, the 7Li resonance reports on the average Li+ structure that is affected by the rapid exchange of TTE and MeCN coordination on the timescale of the NMR experiment. Therefore, as the TTE content is increased, the equilibrium governing TTE vs. MeCN coordination begins to favor TTE coordination resulting in a dynamic shift of the 7Li resonance as a function of TTE concentration. Relating the Li solvation structure to Li-S electrochemistry Raman and NMR spectroscopy coupled with AIMD simulations shows that the addition of TTE to the (MeCN)2−LiTFSI solvate perturbs the first solvation shell of Li+. TTE exchanges with coor-

dinated MeCN, resulting in uncoordinated or free MeCN in the electrolyte solution (Figure 8). We hypothesize the formation of free MeCN in the electrolyte will affect the S electrochemical reduction processes by enhancing the polysulfide solvating capability of the electrolyte. In the absence of polysulfide solvating molecules, the S reduction pathway would be forced to follow a solid-state conversion reaction in lieu of forming soluble polysulfide intermediates. Here, however, the formation of free MeCN as a result of TTE addition allows for the formation of soluble intermediates that facilitate reduction. We observe the enhancement of the formation rate of the S3•− in the (MeCN)x−LiTFSI:TTE electrolyte containing the highest free MeCN content. Interestingly, although the S reduction kinetics are improved likely as a result of free MeCN content, the stability of polysulfides at the cathode surface is unaffected. The persistence of polysulfide at the cathode surface could be a consequence of either limited solubility or the higher viscosity in this electrolyte. The viscosity of the electrolytes with TTE at a volume ratio of 1:1 (8.6 cP)22 is similar to that of the electrolyte that enables fast diffusion, 1 M LiTFSI in DOL/TEGDME (3 cP)43 suggesting that the limited solubility of polysulfides in the solvate electrolytes is the cause for decreased diffusion. Therefore, it is likely that MeCN is able to solubilize the polysulfides at the electrolyte/S interface only thereby maintaining the intermediates near the electrode surface. This behavior is likely to result in good cycling capability as the loss of active material is decreased. The formation of free MeCN could have

Figure 8. Snapshots from thermally equilibrated solution phase (MeCN)2‒LiTFSI and (MeCN)2‒LiTFSI:TTE complexes obtained from AIMD. A fraction of the TTE coordinates to the Li+ via Li‒F interactions at the expense of MeCN coordination. implications on other battery performance metrics such as rate capability and S8 utilization efficiency. However, it is important to mention that free MeCN could also affect the Li deposition and dissolution reactions at the anode of the Li−S battery. Because of these competing effects, it is difficult to deconvolute the possible benefits of free MeCN at the cathode with the possible detrimental effects at the anode. However, a systematic study showing the effect of TTE addition to various solvate complexes on battery performance would be useful. Additionally, it is interesting to note that in the tetraglyme−LiTFSI solvate system, addition of hydrofluoroether to the solvate structure decreases the solubility of polysulfides.16 This is likely due to the stronger binding of the tetraglyme to Li+ as a result of the chelate effect.20 Strong tetraglyme coordination prevents the hydrofluoroether from displacing the bound solvent from the solvate structure.19 Therefore, both the choice of solvent that forms the solvate structure and the choice of cosolvent used to dilute the solvate plays an important role in the behavior of the electrolyte with respect to the cycling of the Li-S cell. We suggest that both the solvent and cosolvent can be tuned to promote some solvent mediated S reduction while still maintaining good

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

electrical contact of the active material at the cathode and preventing dissolution. CONCLUSION The electrochemical behavior of the Li‒S cell using the (MeCN)2‒ LiTFSI solvate electrolyte with a hydrofluoroether, TTE, cosolvent was evaluated with CV and in situ Raman spectroscopy. The CV obtained in (MeCN)2‒LiTFSI with and without TTE is different compared to the conventional electrolyte, 1 M LiTFSI in DOL/TEGDME, suggesting changes in the S reduction and oxidation processes. Using in situ Raman spectroscopy, we observe poor S reduction kinetics in the (MeCN)2‒LiTFSI solvate electrolyte both with and without TTE, despite the low viscosity (MeCN)2‒ LiTFSI:TTE. Diffusion of the polysulfides away from the cathode surface, however, is not observed suggesting greater stability of the polysulfides at the cathode. A similar electrolyte with the same TTE content but slightly less LiTFSI, (MeCN)2.5‒LiTFSI:TTE, enables S reduction kinetics that closely resemble those in the conventional electrolyte while maintaining the enhanced stability of polysulfide at the cathode. Raman spectroscopy and NMR measurements coupled with AIMD simulations show that the addition of TTE to the (MeCN)2,2.5‒LiTFSI electrolytes modifies the Li+ solvation structure, contrary to the previous suggestions that TTE is inert with respect to the solvate. The TTE competes with MeCN coordination in the solvate structure resulting in a higher free MeCN content as TTE is added. TTE exchanges with MeCN more readily in the (MeCN)2.5‒ LiTFSI electrolyte resulting in higher free MeCN content. The content of free MeCN likely facilitates polysulfide formation kinetics through enhanced local solvation effects. Because of the strong effect of the local Li+ solvation structure in the electrolyte on the S reduction kinetics, we suggest that the solvent used to form the solvate and the cosolvent used to dilute the solvate can be tuned to achieve beneficial cycling performance.

discussions. K. C. L. and L. A. C. acknowledge grants of computer time through IBM BlueGene/Q computer through Argonne Leadership Computing Facility (ALCF) and the LCRC Blues Cluster at Argonne National Laboratory. We thank Professor Scott E. Denmark and Guanqun Zhang of the Department of Chemistry at UIUC for the Karl Fisher titration measurements.

REFERENCES (1)

(2)

(3)

(4)

(5)

(6)

(7)

(8)

ASSOCIATED CONTENT Supporting Information Li and 19F NMR of the wet (>50 ppm water) solvate electrolyte before and after Li metal exposure; CVs of the solvate electrolyte neat and diluted with TTE; Raman spectra of the LiTFSI concentration series in MeCN; assignments of Raman modes for the LiTFSI concentration series; Gaussian fits of the ex situ Raman spectra; DFT calculated 7Li NMR shifts of dilute LiTFSI in MeCN, (MeCN)2‒LiTFSI, and (MeCN)2‒LiTFSI:TTE. This material is available free of charge via the Internet at http://pubs.acs.org.

7

(9)

(10)

(11)

AUTHOR INFORMATION (12)

Corresponding Author *[email protected]

(13)

Notes The authors declare no competing financial interests.

ACKNOWLEDGMENTS

(14)

This work was supported as part of the Joint Center for Energy Storage Research, an Energy Innovation Hub funded by the U.S. Department of Energy, Office of Science, Basic Energy Sciences. K. A. S acknowledges post-doctoral funding from the St. Elmo Brady Future Faculty Fellowship. The authors thank Lingyang Zhu for assistance with T1 measurements and Paul M. Bayley for helpful

Page 10 of 13

(15)

Bruce, P. G.; Freunberger, S. A.; Hardwick, L. J.; Tarascon, J.-M. Li-O2 and Li-S Batteries with High Energy Storage. Nat. Mater. 2012, 11, 19–29. Tarascon, J.-M.; Armand, M. Issues and Challenges Facing Rechargeable Lithium Batteries. Nature 2001, 414, 359– 367. Yamin, H.; Gorenshtein, A.; Penciner, J.; Sternberg, Y.; Peled, E. Lithium Sulfur Battery Oxidation/Reduction Mechanisms of Polysulfides in THF Solutions. J. Electrochem. Soc. 1988, 135, 1045–1048. Mikhaylik, Y. V.; Akridge, J. R. Polysulfide Shuttle Study in the Li/S Battery System. J. Electrochem. Soc. 2004, 151, A1969–A1976. Ji, X.; Lee, K. T.; Nazar, L. F. A Highly Ordered Nanostructured Carbon-Sulphur Cathode for LithiumSulphur Batteries. Nat. Mater. 2009, 8, 500–506. Wei Seh, Z.; Li, W.; Cha, J. J.; Zheng, G.; Yang, Y.; McDowell, M. T.; Hsu, P.-C.; Cui, Y. Sulphur–TiO2 Yolk– shell Nanoarchitecture with Internal Void Space for LongCycle Lithium–sulphur Batteries. Nat. Commun. 2013, 4, 1331. Chung, S.-H.; Manthiram, A. Bifunctional Separator with a Light-Weight Carbon-Coating for Dynamically and Statically Stable Lithium-Sulfur Batteries. Adv. Funct. Mater. 2014, 24, 5299–5306. Jin, Z.; Xie, K.; Hong, X.; Hu, Z.; Liu, X. Application of Lithiated Nafion Ionomer Film as Functional Separator for Lithium Sulfur Cells. J. Power Sources 2012, 218, 163– 167. Aurbach, D.; Pollak, E.; Elazari, R.; Salitra, G.; Kelley, C. S.; Affinito, J. On the Surface Chemical Aspects of Very High Energy Density, Rechargeable Li–Sulfur Batteries. J. Electrochem. Soc. 2009, 156, A694–A702. Wu, H.-L.; Huff, L. A.; Gewirth, A. A. In-Situ Raman Spectroscopy of Sulfur Speciation in Lithium-Sulfur Batteries. ACS Appl. Mater. Interfaces 2015, 7, 1709–1719. Mikhaylik, Y. V. Electrolytes for Lithium Sulfur Cells. US7354680 B2, April 8, 2008. Yamada, Y.; Yamada, A. Review—Superconcentrated Electrolytes for Lithium Batteries. J. Electrochem. Soc. 2015, 162, A2406–A2423. Cheng, L.; Curtiss, L. A.; Zavadil, K. R.; Gewirth, A. A.; Shao, Y.; Gallagher, K. G. Sparingly Solvating Electrolytes for High Energy Density Lithium–Sulfur Batteries. ACS Energy Lett 2016, 1, 503–509. Shin, E. S.; Kim, K.; Oh, S. H.; Cho, W. I. Polysulfide Dissolution Control: The Common Ion Effect. Chem. Commun. 2013, 49, 2004–2006. Suo, L.; Hu, Y.-S.; Li, H.; Armand, M.; Chen, L. A New Class of Solvent-in-Salt Electrolyte for High-Energy Rechargeable Metallic Lithium Batteries. Nat. Commun. 2013, 4, 1481.

ACS Paragon Plus Environment

Page 11 of 13 (16)

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

(17)

(18)

(19)

(20)

(21)

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

ACS Applied Materials & Interfaces Dokko, K.; Tachikawa, N.; Yamauchi, K.; Tsuchiya, M.; Yamazaki, A.; Takashima, E.; Park, J.-W.; Ueno, K.; Seki, S.; Serizawa, N.; Watanabe, M. Solvate Ionic Liquid Electrolyte for Li–S Batteries. J. Electrochem. Soc. 2013, 160, A1304–A1310. Ueno, K.; Yoshida, K.; Tsuchiya, M.; Tachikawa, N.; Dokko, K.; Watanabe, M. Glyme–Lithium Salt Equimolar Molten Mixtures: Concentrated Solutions or Solvate Ionic Liquids? J. Phys. Chem. B 2012, 116, 11323–11331. Ueno, K.; Tatara, R.; Tsuzuki, S.; Saito, S.; Doi, H.; Yoshida, K.; Mandai, T.; Matsugami, M.; Umebayashi, Y.; Dokko, K.; Watanabe, M. Li+ Solvation in glyme–Li Salt Solvate Ionic Liquids. Phys. Chem. Chem. Phys. 2015, 17, 8248–8257. Saito, S.; Watanabe, H.; Ueno, K.; Mandai, T.; Seki, S.; Tsuzuki, S.; Kameda, Y.; Dokko, K.; Watanabe, M.; Umebayashi, Y. Li+ Local Structure in Hydrofluoroether Diluted Li-Glyme Solvate Ionic Liquid. J. Phys. Chem. B 2016, 120, 3378–3387. Zhang, C.; Yamazaki, A.; Murai, J.; Park, J.-W.; Mandai, T.; Ueno, K.; Dokko, K.; Watanabe, M. Chelate Effects in Glyme/Lithium Bis(trifluoromethanesulfonyl)amide Solvate Ionic Liquids, Part 2: Importance of Solvate-Structure Stability for Electrolytes of Lithium Batteries. J. Phys. Chem. C 2014, 118, 17362–17373. See, K. A.; Leskes, M.; Griffin, J. M.; Britto, S.; Matthews, P. D.; Emly, A.; Van der Ven, A.; Wright, D. S.; Morris, A. J.; Grey, C. P.; Seshadri, R. Ab Initio Structure Search and in Situ 7Li NMR Studies of Discharge Products in the Li–S Battery System. J. Am. Chem. Soc. 2014, 136, 16368– 16377. Cuisinier, M.; Cabelguen, P.-E.; Adams, B. D.; Garsuch, A.; Balasubramanian, M.; Nazar, L. F. Unique Behaviour of Nonsolvents for Polysulphides in Lithium–sulphur Batteries. Energy Environ. Sci. 2014, 7, 2697. Han, S.-D.; Borodin, O.; Seo, D. M.; Zhou, Z.-B.; Henderson, W. A. Electrolyte Solvation and Ionic Association V. Acetonitrile-Lithium Bis(fluorosulfonyl)imide (LiFSI) Mixtures. J. Electrochem. Soc. 2014, 161, A2042–A2053. Dey, A. N.; Holmes, R. W. Safety Studies on Li / SO2 Cells I . Differential Thermal Analysis (DTA) of Cell Constituents. J. Electrochem. Soc. 1979, 126, 1637–1644. Yamada, Y.; Furukawa, K.; Sodeyama, K.; Kikuchi, K.; Yaegashi, M.; Tateyama, Y.; Yamada, A. Unusual Stability of Acetonitrile-Based Superconcentrated Electrolytes for Fast-Charging Lithium-Ion Batteries. J. Am. Chem. Soc. 2014, 136, 5039-5046. Sodeyama, K.; Yamada, Y.; Aikawa, K.; Yamada, A.; Tateyama, Y. Sacrificial Anion Reduction Mechanism for Electrochemical Stability Improvement in Highly Concentrated Li-Salt Electrolyte. J. Phys. Chem. C 2014, 118, 14091–14097. Lu, H.; Yuan, Y.; Zhang, K.; Qin, F.; Lai, Y.; Liu, Y. Application of Partially Fluorinated Ether for Improving Performance of Lithium/Sulfur Batteries. J. Electrochem. Soc. 2015, 162, A1460–A1465. Choi, J.-W.; Kim, J.-K.; Cheruvally, G.; Ahn, J.-H.; Ahn, H.-J.; Kim, K.-W. Rechargeable Lithium/sulfur Battery with Suitable Mixed Liquid Electrolytes. Electrochimica Acta 2007, 52, 2075–2082. Zu, C.; Azimi, N.; Zhang, Z.; Manthiram, A. Insight into Lithium–metal Anodes in Lithium–sulfur Batteries with a

(30)

(31)

(32)

(33)

(34)

(35) (36)

(37)

(38)

(39) (40)

(41)

(42)

Fluorinated Ether Electrolyte. J. Mater. Chem. A 2015, 3, 14864–14870. Azimi, N.; Xue, Z.; Bloom, I.; Gordin, M. L.; Wang, D.; Daniel, T.; Takoudis, C.; Zhang, Z. Understanding the Effect of a Fluorinated Ether on the Performance of Lithium–Sulfur Batteries. ACS Appl. Mater. Interfaces 2015, 7, 9169–9177. Azimi, N.; Weng, W.; Takoudis, C.; Zhang, Z. Improved Performance of Lithium–sulfur Battery with Fluorinated Electrolyte. Electrochem. Commun. 2013, 37, 96–99. Gordin, M. L.; Dai, F.; Chen, S.; Xu, T.; Song, J.; Tang, D.; Azimi, N.; Zhang, Z.; Wang, D. Bis(2,2,2Trifluoroethyl) Ether As an Electrolyte Co-Solvent for Mitigating Self-Discharge in Lithium–Sulfur Batteries. ACS Appl. Mater. Interfaces 2014, 6, 8006–8010. Wu, H.-L.; Huff, L. A.; Esbenshade, J. L.; Gewirth, A. A. In Situ EQCM Study Examining Irreversible Changes the Sulfur–Carbon Cathode in Lithium–Sulfur Batteries. ACS Appl. Mater. Interfaces 2015, 7, 20820–20828. Long, B. R.; Chan, M. K. Y.; Greeley, J. P.; Gewirth, A. A. Dopant Modulated Li Insertion in Si for Battery Anodes: Theory and Experiment. J. Phys. Chem. C 2011, 115, 18916–18921. CPMD V4.1 Copyright IBM Corp 1990-2015, Copyright MPI fuer Festkoerperforschung Stuttgart 1997-2001. Perdew; Burke; Ernzerhof. Generalized Gradient Approximation Made Simple. Phys. Rev. Lett. 1996, 77 , 3865– 3868. Grimme, S.; Ehrlich, S.; Goerigk, L. Effect of the Damping Function in Dispersion Corrected Density Functional Theory. J. Comput. Chem. 2011, 32, 1456–1465. Bylander, D. M.; Kleinman, L. Energy Fluctuations Induced by the Nose Thermostat. Phys. Rev. B 1992, 46, 13756–13761. Hoover, W. G. Canonical Dynamics: Equilibrium PhaseSpace Distributions. Phys. Rev. A 1985, 31, 1695–1697. Gaussian 09, Revision E.01, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji, H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A., Jr.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian, Inc., Wallingford CT, 2009. Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994, 98, 11623–11627. Barchasz, C.; Molton, F.; Duboc, C.; Leprêtre, J.-C.; Patoux, S.; Alloin, F. Lithium/Sulfur Cell Discharge Mechanism: An Original Approach for Intermediate Species Identification. Anal. Chem. 2012, 84, 3973–3980.

ACS Paragon Plus Environment

ACS Applied Materials & Interfaces (43)

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

(44)

(45)

(46)

(47)

(48)

(49)

(50)

(51)

(52)

(53)

(54)

Kim, H. S.; Jeong, T.-G.; Choi, N.-S.; Kim, Y.-T. The Cycling Performances of Lithium–sulfur Batteries in TEGDME/DOL Containing LiNO3 Additive. Ionics 2013, 19, 1795–1802. Xu, R.; Lu, J.; Amine, K. Progress in Mechanistic Understanding and Characterization Techniques of Li-S Batteries. Adv. Energy Mater. 2015, 5, 1500408. Lu, Y.-C.; He, Q.; Gasteiger, H. A. Probing the Lithium– Sulfur Redox Reactions: A Rotating-Ring Disk Electrode Study. J. Phys. Chem. C 2014, 118, 5733–5741. Martin, R. P.; Doub, W. H.; Roberts, J. L.; Sawyer, D. T. Electrochemical Reduction of Sulfur in Aprotic Solvents. Inorg. Chem. 1973, 12, 1921–1925. Yeon, J.-T.; Jang, J.-Y.; Han, J.-G.; Cho, J.; Lee, K. T.; Choi, N.-S. Raman Spectroscopic and X-Ray Diffraction Studies of Sulfur Composite Electrodes during Discharge and Charge. J. Electrochem. Soc. 2012, 159, A1308–A1314. Fujinaga, T.; Kuwamoto, T.; Okazaki, S.; Hojo, M. Electrochemical Reduction of Elemental Sulfur in Acetonitrile. Bull. Chem. Soc. Jpn. 1980, 53, 2851–2855. Bakker, A.; Gejji, S.; Lindgren, J.; Hermansson, K.; Probst, M. M. Contact Ion Pair Formation and Ether Oxygen Coordination in the Polymer Electrolytes M[N(CF3SO2)2]2PEOn for M = Mg, Ca, Sr and Ba. Polymer 1995, 36, 4371–4378. Edman, L. Ion Association and Ion Solvation Effects at the Crystalline−Amorphous Phase Transition in PEO−LiTFSI. J. Phys. Chem. B 2000, 104, 7254–7258. Herstedt, M.; Smirnov, M.; Johansson, P.; Chami, M.; Grondin, J.; Servant, L.; Lassègues, J. C. Spectroscopic Characterization of the Conformational States of the Bis(trifluoromethanesulfonyl)imide Anion (TFSI−). J. Raman Spectrosc. 2005, 36, 762–770. Brouillette, D.; Irish, D. E.; Taylor, N. J.; Perron, G.; Odziemkowski, M.; Desnoyers, J. E. Stable Solvates in Solution of Lithium Bis(trifluoromethylsulfone)imide in Glymes and Other Aprotic Solvents: Phase Diagrams, Crystallography and Raman Spectroscopy. Phys. Chem. Chem. Phys. 2002, 4, 6063–6071. Seo, D. M.; Borodin, O.; Han, S.-D.; Boyle, P. D.; Henderson, W. A. Electrolyte Solvation and Ionic Association II. Acetonitrile-Lithium Salt Mixtures: Highly Dissociated Salts. J. Electrochem. Soc. 2012, 159, A1489–A1500. Neelakantan, P. Raman Spectrum of Acetonitrile. Proc. Indian Acad. Sci. - Sect. A 1964, 60, 422–424.

(55)

(56)

(57)

(58)

(59)

(60)

(61)

(62)

(63)

(64)

Page 12 of 13

Fini, G.; Mirone, P. On the Secondary Structure of Some Vibrational Bands of Acetonitrile. Spectrochim. Acta Part Mol. Spectrosc. 1976, 32, 439–440. Barthel, J.; Buchner, R.; Wismeth, E. FTIR Spectroscopy of Ion Solvation of LiClO4 and LiSCN in Acetonitrile, Benzonitrile, and Propylene Carbonate. J. Solut. Chem. 2000, 29, 937–954. Brouillette, D.; Irish, D. E.; Taylor, N. J.; Perron, G.; Odziemkowski, M.; Desnoyers, J. E. Stable Solvates in Solution of Lithium Bis(trifluoromethylsulfone)imide in Glymes and Other Aprotic Solvents: Phase Diagrams, Crystallography and Raman Spectroscopy. Phys. Chem. Chem. Phys. 2002, 4, 6063–6071. Dallek, S.; James, S. D.; Kilroy, W. P. Exothermic Reactions among Components of Lithium‐Sulfur Dioxide and Lithium‐Thionyl Chloride Cells. J. Electrochem. Soc. 1981, 128, 508–516. Cahen, Y. M.; Handy, P. R.; Roach, E. T.; Popov, A. I. Spectroscopic Studies of Ionic Solvation. XVI. Lithium-7 and Chlorine-35 Nuclear Magnetic Resonance Studies in Various Solvents. J. Phys. Chem. 1975, 79, 80–85. Bloembergen, N.; Purcell, E. M.; Pound, R. V. Relaxation Effects in Nuclear Magnetic Resonance Absorption. Phys. Rev. 1948, 73, 679–712. Craig, R. A.; Richards, R. E. Nuclear Magnetic Resonance in Lithium Chloride Solutions. Trans. Faraday Soc. 1963, 59, 1972–1982. Bayley, P. M.; Best, A. S.; MacFarlane, D. R.; Forsyth, M. The Effect of Coordinating and Non-Coordinating Additives on the Transport Properties in Ionic Liquid Electrolytes for Lithium Batteries. Phys. Chem. Chem. Phys. 2011, 13, 4632. Richardson, P. M.; Voice, A. M.; Ward, I. M. NMR T1 Relaxation Time Measurements and Calculations with Translational and Rotational Components for Liquid Electrolytes Containing LiBF4 and Propylene Carbonate. J. Chem. Phys. 2013, 139, 214501. Ali, F.; Forsyth, M.; Garcia, M. C.; Smith, M. E.; Strange, J. H. A 7Li and 19F NMR Relaxation Study of LiCF3SO3 in Plasticised Solid Polyether Electrolytes. Solid State Nucl. Magn. Reson. 1995, 5, 217–225.

ACS Paragon Plus Environment

Page 13 of 13

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

ACS Applied Materials & Interfaces

For Table of Contents Only

ACS Paragon Plus Environment