
==== Front
J Phys Chem C Nanomater Interfaces
J Phys Chem C Nanomater Interfaces
jy
jpccck
The Journal of Physical Chemistry. C, Nanomaterials and Interfaces
1932-7447
1932-7455
American Chemical Society

10.1021/acs.jpcc.4c03412
Article
Development of Fluoride-Ion Primary Batteries: The Electrochemical Defluorination of CFx
https://orcid.org/0009-0007-6991-9401
Robinson Loleth E. †
Wang Jonah †
Asare Harrison †‡§
Andrews Jessica L. ∥
Tripathi Balram #
Katiyar Ram #
https://orcid.org/0000-0002-7078-8206
Melot Brent C. ∥⊥
https://orcid.org/0000-0002-5537-3870
Messinger Robert J. *†§
Jones Simon C. *∇
West William C. *○
† Department of Chemical Engineering, The City College of New York, CUNY, New York, New York 10031, United States
‡ Department of Chemistry and Biochemistry, The City College of New York, CUNY, New York, New York 10031, United States
§ Ph.D. Program in Chemistry, The Graduate Center of the City University of New York, New York, New York 10016, United States
∥ Department of Chemistry, University of Southern California, Los Angeles, California 90089, United States
⊥ Department of Chemical Engineering and Materials Science, University of Southern California, Los Angeles, California 90089, United States
# Department of Physics and Institute for Functional Nanomaterials, University of Puerto Rico, San Juan 00925-2537, Puerto Rico
∇ Flion Energy Inc., Pasadena, California 91107, United States
○ Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California 91109, United States
* Email: rmessinger@ccny.cuny.edu.
* Email: simon.jones@flionenergy.com.
* Email: william.c.west@jpl.nasa.gov.
15 08 2024
29 08 2024
128 34 1419514205
23 05 2024
08 08 2024
06 08 2024
© 2024 The Authors. Published by American Chemical Society
2024
The Authors
https://creativecommons.org/licenses/by/4.0/ Permits the broadest form of re-use including for commercial purposes, provided that author attribution and integrity are maintained (https://creativecommons.org/licenses/by/4.0/).

The lithium–carbon monofluoride (Li-CFx) couple has the highest specific energy of any practical battery chemistry. However, the large polarization associated with the CFx electrode (>1.5 V loss) limits it from achieving its full discharge energy, motivating the search for new CFx reaction mechanisms with reduced overpotential. Here, using a liquid fluoride (F)-ion conducting electrolyte at room temperature, we demonstrate for the first time the electrochemical defluorination of CFx cathodes, where metal fluorides form at a metal anode instead of the CFx cathode. F-ion primary cells were developed by pairing CFx cathodes with either lead (Pb) or tin (Sn) metal anodes, which achieved specific capacities of over 700 mAh g–1 and over 400 mAh g–1, respectively. Solid-state 19F and 119Sn{19F} nuclear magnetic resonance (NMR), X-ray diffraction (XRD), Raman, inductively coupled plasma (ICP), and X-ray fluorescence (XRF) measurements establish that upon discharge, the CFx cathode defluorinates while Pb forms PbF2 and Sn forms both SnF4 and SnF2. Technological development of F-ion metal-CFx cells based on this concept represents a promising avenue for realizing primary batteries with high specific energy.

National Science Foundation 10.13039/100000001 DGE-1842487 Flion Energy Inc NA NA National Aeronautics and Space Administration 10.13039/100000104 80NSSC19M0236 National Aeronautics and Space Administration 10.13039/100000104 80NSSC19M0199 National Aeronautics and Space Administration 10.13039/100000104 80NM0018D0004 document-id-old-9jp4c03412
document-id-new-14jp4c03412
ccc-price
==== Body
pmcIntroduction

The lithium–carbon monofluoride (Li-CFx) primary (nonrechargeable) battery cell is employed across a broad spectrum of commercial uses, from implantable medical devices to marine, military, medical, and space applications.1−4 This battery chemistry offers the highest specific energy (∼2180 Wh kg–1, based on a theoretical discharge potential of 4.57 V)1 and specific capacity (865 mAh g–1, based on CFx cathode when x = 1) among all lithium nonrechargeable chemistries commercially available.2 In many cases, the limited capacity of power sources is the key performance bottleneck for various technologies that underpin modern technological advancements; therefore, greater energy is always desirable. For example, the Li-CFx chemistry is of significant interest for future spacecraft applications, especially where a photovoltaic or radioisotope power is either unavailable or impractical.5,6 As applications become more complex and more energy-demanding, achieving higher practical energy output for longer use times is critical, highlighting the need to identify new cell chemistries.

The electrochemical conversion mechanism for the conventional Li-CFx cell discharge process is (1) the anodic oxidation of the lithium metal anode, Li → Li+ + e–, (2) the transport of lithium ions through the electrolyte, and (3) the cathodic reduction of the CFx to form carbon (C) and lithium fluoride (LiF) with the discharge products remaining at the cathode, CFx + xLi → (LiF)x + C. The overall spontaneous electrochemical conversion reaction is CFx + xLi+ + xe– → (LiF)x + C.

Although the theoretical discharge voltage of the Li-CFx cell is >4 V, the practical discharge voltage (∼2.5 V), even at low current densities, is significantly lower than the theoretical value. This ca. 2 V polarization loss represents a massive penalty in specific energy relative to the theoretical value. For example, a Li-CFx D-cell discharged at 2.5 V offers a specific energy of 700 Wh kg–1. If the same cell was discharged at 4.5 V, then the specific energy achieved is closer to 1260 Wh kg–1, a 560 Wh kg–1 (or 80%) practical gain. Most importantly, the higher energy output provided by a 4.5 V discharge voltage design could replace other primary cell chemistries such as lithium-thionyl chloride (Li-SOCl2) without the need to change the power electronics design. The low practical discharge voltage of Li-CFx has been ascribed to several mechanisms.7−12 Whittingham9 proposed a two-phase reaction mechanism, where the discharge product is a nonstoichiometric ternary intercalation compound (CLixF) that later decomposes into carbon and LiF. Nonstoichiometric solvent-mediated ternary compounds have also been proposed (CFLix:Sy), where co-intercalated Li+ and solvent molecules later decompose to the final discharge products.10,12 Recently, Leung et al.8 proposed a solvent-mediated edge propagation mechanism and demonstrated a correlation between the solvent-based lithium intercalation complex and the CFx operating voltage window using density functional theory methods.

We hypothesize that by fundamentally changing the discharge process of the Li-CFx cell chemistry, it may be possible to access a greater discharge potential. One possible route to do so is to change the discharge process from lithiating the CFx cathode to defluorinating the CFx cathode and fluorinating the anode. Clearly, the discharge mechanism and location of the discharge product (metal fluorides) are entirely different in this cell relative to conventional metal anode −CFx cells, which raises the possibility of improvement to a higher practical discharge voltage. The reversible fluorination of CuF2 and BiF3 has been demonstrated previously; however, applications are limited since high temperatures (∼150 °C) are necessary to cycle the battery at practical rates using a solid F-ion conductor.13,14 Recent reports describe the development of a room temperature fluoride-ion conducting electrolyte.15,16 However, a compatible metal anode is necessary, given the strong reactivity between the lithium metal and the F-ion electrolyte.

In this work, using a liquid F-ion conducting electrolyte, we demonstrate the electrochemical defluorination of CFx electrodes at room temperature versus either lead Pb or Sn metal anodes. Solid-state 19F and 119Sn{19F} nuclear magnetic resonance (NMR), X-ray diffraction (XRD), inductively coupled plasma (ICP), Raman, and X-ray fluorescence (XRF) measurements evidence the defluorination of the CFx electrode upon discharge and concurrent formation of PbF2 at the Pb anode or SnF4 and SnF2 at the Sn anode. XRD, ICP, and XRF revealed that only modest transport of either Pb or Sn between electrodes occurred; however, cation transport is reduced substantially when using Sn. To the best of our knowledge, this is the first report of the electrochemical defluorination of a CFx electrode, which is reduced to carbon, with the concomitant fluorination of a metal anode as the conjugate electrochemical reaction.

Methods

Electrode and Cell Preparation

The CFx cathodes were prepared in-house by mixing 92 wt % CFx powder (ARC, x ≈ 1, theoretical capacity of 865 mAh g–1), 5 wt % carbon black (Super P, Alfa Aesar), and 3 wt % poly(vinylidene fluoride) (PVDF; Aldrich) dissolved in N-methyl-2-pyrrolidone (NMP; Aldrich) to prepare a slurry. Note that x ≈ 1 in CFx represents the average composition of the bulk active material; however, local differences in fluorination are expected, giving rise to compositional heterogeneity within the CFx structure. Self-standing Pb metal electrodes were prepared using a composition of 80 wt % Pb metal powder (CERAC Inc., 200 mesh), 10 wt % carbon black, and 10 wt % poly(tetrafluoroethylene) (PTFE; Aldrich). The Pb slurry was prepared in a mortar and pestle using acetone as a solvent and later calendared using a manual laboratory roller to form a self-standing electrode. The Sn metal electrodes were also prepared using a slurry composed of 80 wt % Sn metal powder (CERAC Inc.), 10 wt % carbon black, and 10 wt % PVDF dissolved in N-methylpyrrolidone (NMP). Prior to casting, the Sn electrode powdered components were ball milled (SPEX 8000) for 2 h. The CFx and Sn slurries were cast on a 0.127 mm thick stainless steel foil (Shop-Aid, Inc.) precoated with a layer of carbon ink (EB-012 Henkel Corp). All electrodes were dried overnight at 100 °C under vacuum prior to cell assembly.

The composite CFx cathode, Sn anode, and Pb anode mass loadings were 5.93, 23, and 248 mg cm–2, respectively. Stainless steel CR2032 coin cell cases were stripped from their interior aluminum coating to prevent undesired reactions with the electrolyte solvent. To remove the aluminum coating from the coin cell casings, they were placed in a 1 M solution of NaOH (J.T. Baker) for 20 min; then, they were rinsed with deionized water and dried overnight. Coin cells were prepared using dried CFx and Sn metal electrodes with 16 mm diameter, Pb metal electrodes with 13 mm diameter, separators (Celgard 2325) with 17 mm diameter, and 100 μL of electrolyte consisting of 0.75 M N,N,N-trimethyl-N-neopentylammonium fluoride (Np1F) dissolved in bis(2,2,2-trifluoroethyl) ether (BTFE; Aldrich), under an argon atmosphere (H2O and O2 < 1 ppm). The Np1F fluoride salt was synthesized as described in detail by Davis et al.15 Following manual crimp sealing, all coin cells were galvanostatically discharged using an Arbin LBT battery tester at 10 μA discharge current to a cutoff voltage of 0 V. All metal-CFx cells were designed with anode specific capacities far in excess of the cathode specific capacities; thus, they are cathode-limited in capacity. All specific capacities are reported per mass of CFx (865 mAh g–1, based on the CFx cathode when x ≈ 1). Calculations of the specific capacity are detailed in Text S1, Supporting Information.

X-ray Diffraction

Powder XRD measurements were carried out using an Aeris PANalytical X-Pert system with a diffractometer run in the θ – 2θ geometry, with a Cu anode (λ = 1.541 A) at an accelerating voltage of 40 kV and a tube current of 15 mA, standard phase ID measurement profile measured from 10 to 100° 2θ using a divergence slit = 0.125°, Soller slit = 0.02 rad, with step size = 0.01°.

The resulting XRD patterns for both the pristine and discharged Pb and Sn electrodes were refined against published structures using the Rietveld method as implemented in the TOPAS-Academic suite,17 with the resulting structural parameters given in Tables S1–S4.

For XRD sample preparation, CFx, Sn, and Pb electrodes were harvested from discharged coin cells inside a dry room with H2O content <5 ppm. Ethanol (Fischer Scientific) was used as a rinsing solvent to remove residual organic electrolyte from the harvested electrodes prior to measurements. Pristine CFx and Sn electrodes were used as prepared. The interlayer distance extracted from the diffraction pattern of the pristine CFx electrode is 0.7 nm.

Scanning Electron Microscopy (SEM)

SEM was performed by using a Zeiss Supra 55 field emission scanning electron microscope under high vacuum. The pristine CFx electrode was gold plated prior to imaging to prevent surface charging.

Inductively Coupled Plasma-Optical Emission Spectroscopy (ICP-OES)

Inductively coupled plasma-optical emission spectroscopy (ICP-OES) was carried out using a PerkinElmer Optima 7300 DV ICP operating with an argon plasma flow of 15 and 0.2 L min–1 auxiliary flow. Nebulizer flow was 0.55 L min–1 with a peristaltic pump sample introduction flow rate of 1.5 mL min–1. Radio frequency (rf) power was 1450 W, and the instrument optics was set for an axial view of the plasma. The ICP was calibrated using a top-level standard containing Sn and Pb, present at 1.0 ppm (1.0 μg mL–1), in 2% HNO3 (prepared from Ultrex ultrapure nitric acid and deionized water). For standards and samples, triplicate runs were performed, and the results were averaged. The electrodes were placed into ICP vials containing ∼5 g of ASTM type I water and allowed to stand overnight. Samples were analyzed as-is, along with an ASTM type I water sample as a blank.

Raman Spectroscopy

Raman spectra were measured using the excitation wavelength of 514.5 nm of an argon ion laser at a power of 1 mW.

X-ray Fluorescence Spectroscopy

XRF measurements were carried out using a Horiba model XGT-9000 X-ray fluorescence microscope (μXRF), using the following settings: live time: 100 s; processing time: P2; XGT diameter: 50 μm; X-ray tube voltage: 30 kV; current: automatic. Samples were scanned at two different locations with a 1.2 mm beam size and averaged to increase the area response.

Solid-State Nuclear Magnetic Resonance Spectroscopy

Solid-state NMR experiments were acquired on a Bruker Avance III HD 600 NMR spectrometer with a 14.1 T narrow-bore superconducting magnet operating at 564.69 and 223.79 MHz for 19F and 119Sn nuclei, respectively. A Phoenix NMR 1.6 mm HXY magic-angle-spinning (MAS) probehead was used, where all measurements were conducted with MAS rates of 40 kHz unless specified otherwise. Air was pumped through the probehead at 293.2 K and 600 L h–1 to mitigate sample heating due to MAS. 19F chemical shifts were referenced to CFCl3 at 0 ppm using pristine LiF (Sigma-Aldrich, ≥99.99%) at −204 ppm as a secondary chemical shift reference. 7Li shifts were referenced to 1 M aqueous LiNO3 at 0 ppm using a pristine LiF shift at −1 ppm as a secondary chemical shift reference. 119Sn chemical shifts were referenced indirectly by scaling the 119Sn frequency with respect to 7Li nuclei by using the ratio of their gyromagnetic ratios, γ119Sn/γ7Li (0.965).

Solid-state 19F experiments used radio frequency (rf) field strengths of 147 kHz (π/2 of 1.7 μs) for all broad-band rf pulses. Solid-state 19F spin-echo NMR experiments were acquired to eliminate the 19F probe and rotor background using a half-echo delay of six rotor periods (150 μs). Recycle delays (25 s) were calibrated such that all 19F nuclear spins relaxed to thermal equilibrium (>5 × T1, the longitudinal relaxation time) between scans. For the CFx electrode, the relative populations of the 19F signals of the solid-state 19F spin-echo NMR experiments can be quantitatively analyzed because the total echo delay of 300 μs is smaller than the transverse 19F longitudinal (T2) relaxation times of the different 19F signals (ranged from 1.5 to 6.7 ms). Two-dimensional (2D) 19F{19F} finite pulse radiofrequency-driven recoupling (fp-RFDR) NMR experiments were acquired to probe the molecular-scale proximities of 19F moieties using mixing times of 1.6 ms. The fp-RFDR π pulses were phase cycled according to the (XY-8)41 super cycle, where the ratio of the π pulse duration (1.7 μs) to the rotor period (25 μs) was 0.068.

To enhance 119Sn signal sensitivity, solid-state 119Sn{19F} cross-polarization magic-angle-spinning (CP-MAS) experiments were used to characterize 119Sn nuclei. 19F and 119Sn rf field strengths of 100 and 60 kHz were used, corresponding to the zero-quantum n = +1 Hartmann–Hahn matching condition. 119Sn{19F} CP-MAS contact times of 4 ms were used for all experiments. The solid-state 2D 119Sn{19F} heteronuclear correlation (HETCOR) NMR experiments are 2D analogues of the one-dimensional (1D) 119Sn{19F} CP-MAS NMR conditions, enhancing the spectral resolution. STATES-TPPI quadrature detection was used in the indirect dimension for all 2D experiments.

For NMR sample preparation, CFx and Sn electrodes were harvested from discharged coin cells inside an argon-filled glovebox (H2O < 1.0 ppm and O2 < 1 ppm). Coin cells were decrimped using an MTI electric (de)crimping machine (model MSk-160E), where the decrimping pressure was set to 2016 psi. The active material was stripped from the stainless steel current collector, crushed using a mortar and pestle, and packed in a 1.6 mm diameter zirconia rotor inside the glovebox. Commercial SnF2 (Thermo Fisher Scientific, 97.5%) and SnF4 (Thermo Fisher Scientific, 99%) were used as purchased and packed in the glovebox. Solid-state NMR spectra were deconvoluted using DMFit.18

Results and Discussion

Electrochemical Performance of F-Ion Sn-CFx and Pb-CFx Primary Cells

To electrochemically defluorinate the CFx cathode, the C–F bonds in the CFx active material must be broken, followed by the F-ions migrating to the anode through a fluoride-ion conducting electrolyte, where they react with the reduced metal ion to form a fluorinated metal discharge product at the anode. The liquid fluoride-ion conducting electrolyte, 0.75 M Np1F dissolved in BTFE, offers a wide electrolyte operating voltage window (4.1 V) and high ionic conductivity (10–3–10–2 S cm–1) that allows for compatibility with a variety of metals.15 We tested the electrochemical discharge performance of the CFx electrode against different metal counter electrodes. A CFx cell against a Pb metal counter electrode was prepared and discharged to 0 V (Figure 1A). The Pb-CFx cell was tested by applying a low initial current of 2.5 μA cm–2, and the current was gradually increased to determine the bounds of the practical current necessary to effectively discharge the cell. Once the current density increased to 12.5 μA cm–2, the characteristic flat discharge plateau of CFx was observed at ca. 0.6 V and achieved a specific capacity of 722.1 mAh g–1.

Figure 1 (A) Galvanostatic discharge of a F-ion Pb-CFx primary cell at the current densities noted. (B) Galvanostatic discharge of F-ion Sn-CFx primary cells discharged at a rate of 5 μA cm–2.

As shown above, ICP measurements of the harvested CFx electrode from the Pb-CFx cell after discharge revealed that the PbF2 discharge product was present at both the anode and cathode, indicating that mass transport of a Pb species is occurring. While the precise lead species is unknown, if the PbF2 was formed due to Pb2+ transport and reaction according to xPb2+ + 2CFx + 2xe– → xPbF2 + 2C at the CFx electrode, we estimate that 5.4% of the discharge capacity was due to Pb2+ ion transport and electrochemical PbF2 formation, with the balance of discharge capacity (94.6%) proceeding by the electrochemical defluorination of CFx and subsequent F-ion transport.

Importantly, the voltage of the F-ion Pb-CFx cell is strongly suggestive that a much higher operating voltage of the CFx cathode can be obtained in this novel cell design if a Li anode were used. For example, the discharge plateau of the F-ion Pb-CFx cell is approximately 0.6 V. Comparing the standard half-cell potentials of Pb (−0.13 V vs standard hydrogen electrode (SHE) for Pb → Pb2+ + 2e–) and Li (−3.01 V vs SHE for Li → Li+ + e–),19 replacing the Pb anode with a Li anode would result in a ca. 3.5 V cell vs Li/Li+ (activities of the reactants and products associated with the different half-cell reactions would result in an additional minor correction). This calculation indicates that the CFx discharge plateau would be closer to the theoretical discharge potential of CFx (e.g., compared to the 2.5–2.7 V typically observed for Li-CFx). Thus, replacing Pb with Li could also allow for a much greater discharge voltage.

Sn metal was chosen in an attempt to find a compatible anode material with reduced cation transport. F-ion Sn-CFx cells were fabricated, using a 3-fold increase in the Sn electrodes mass loading to ensure that the CFx is the capacity-limiting electrode. The electrochemical performance of several Sn-CFx cells was observed by discharging cells under a constant current density of 5 μA cm–2, from the open-circuit voltage (OCV) of ∼1.0 to 0 V (Figure 1B). Specific capacities ranging from 374.2 to 438.7 mAh g–1 were observed and presumed to be dependent on the extent of defluorination of the CFx electrode. A small initial voltage plateau was observed at 0.27 V, followed by the main voltage plateau at approximately 0.15 V. The two discharge plateaus correspond to the formation of two fluorinated discharge products, SnF2 and SnF4, as shown below. Notably, the Sn-CFx cell approached roughly ∼50% of the theoretical capacity of CFx (865 mAh g–1 when x = 1), without optimization, at room temperature.

As shown below, solid-state NMR measurements of the harvested Sn electrode from the Sn-CFx cell revealed SnF4 present at the CFx electrode, which is consistent with XRF measurements. This result indicates that there is mass transport of a Sn species from the anode to the CFx cathode. While the precise tin species is unknown, if the SnF2 formed due to Sn2+ transport to the CFx cathode and reaction according to xSn2+ + 2CFx + 2xe– → xSnF2 + 2C. In principle, Sn could further electrochemically oxidize (Sn2+ →Sn4+ + 2e–) at the tin metal anode, with subsequent electrochemical reduction at the CFx cathode to xSn4+ + 4CFx + 4xe– → xSnF4 + 4C. We estimate that 3.1% of the discharge capacity was achieved by Sn2+ transport, a >40% relative decrease from the Pb2+ case. Importantly, we estimate that electrochemically defluorinating CFx reduces the overall polarization loss by approximately 1.1 and 0.6 V when using Pb and Sn (e.g., compared to approximately 2 V polarization loss observed when lithiating CFx), respectively.

Electrode Crystal Structures and Morphologies

Powder XRD measurements were acquired on a pristine CFx electrode and discharged CFx and Pb electrodes to characterize the structural changes after discharge (Figure 2). The pristine CFx electrode (Figure 2A, gray) exhibits two reflections that correspond to the carbon ink precoating at ca. 2θ of 27 and 54°, as well as two reflections from the stainless steel current collector at ca. 2θ of 44 and 75°. The broad CFx reflection at ca. 2θ of 42° is consistent with the disordered nature of the CFx structure. The XRD pattern of the CFx electrode after discharge against the Pb anode (Figure 2A, black) exhibits multiple new reflections between ca. 2θ of 40 and 79° corresponding to PbF2 in cubic space group Fm3̅m (no. 225). Rietveld refinements of this structure against the XRD data revealed the presence of cubic PbF2 in both the discharged CFx and Pb electrodes, evidencing the transport of Pb ions to the CFx cathode during discharge (Figure 2B,C). The multiphase Rietveld refinement of the discharged Pb anode also indicates the presence of orthorhombic PbF2 (Pnma, no. 62), cubic Pb (Fd3̅m, no. 225), and an orthorhombic PbO impurity (Pbcm, no. 57). The PbO impurity was likely introduced during the electrode casting process, where high surface area Pb metal powder was ground at ambient conditions.

Figure 2 XRD pattern of a (A) discharged CFx electrode from an F-ion Pb-CFx cell (black), pristine CFx electrode (gray), CFx powder (light green), carbon ink (dark green), and stainless steel foil (blue). (B) Rietveld refinement of a CFx electrode from a discharged F-ion Pb-CFx cell. (C) Rietveld refinement of a discharged Pb electrode from a F-ion Pb-CFx cell.

XRD measurements were also collected on a CFx cathode discharged against a Sn anode, as well as the corresponding discharged Sn electrode (Figure 3). The CFx electrode (Figure 3A, black) reveals multiple low-angle reflections associated with the residual electrolyte (Figure S1). Critically, reflections associated with Sn, as observed in the pristine Sn electrode (Figure 3B), or a Sn–F binary phase were not detected in the XRD pattern of the discharged CFx electrode, suggesting that Sn ion transport is minimal during discharge. The XRD data for the discharged Sn electrode (Figure 3C) exhibit reflections corresponding to both tetragonal Sn (I41/amd, no. 141) and monoclinic SnF2 (C12/c1, no. 15).

Figure 3 XRD pattern of (A) discharged CFx electrode from a F-ion Sn-CFx cell (black), pristine CFx electrode (gray), CFx powder (yellow), carbon ink (orange), and stainless steel foil (purple). Rietveld refinement of (B) pristine Sn electrode and (C) discharged Sn electrode from an F-ion Sn-CFx cell.

Morphological changes in the CFx and Sn electrodes were also visualized before and after discharge using SEM (Figure 4). The surface morphology of the CFx electrode becomes considerably smoother upon discharge. Note that in Li-CFx cells, insulating LiF is formed during electrochemical discharge; here, no additional discharge products are observed, consistent with the electrochemical defluorination mechanism proposed here. The composite Sn metal electrode exhibits a relatively inhomogeneous distribution of active material, a result of the difficulty in casting uniform Sn electrodes due to Sn metal being a dense active material.

Figure 4 SEM images of (A) pristine and (B) discharged CFx electrode and (C) pristine and (D) discharged Sn electrode. Both the CFx and Sn discharged electrodes were harvested from an F-ion Sn-CFx cell. The pristine CFx electrode was gold plated prior to imaging to prevent surface charging.

Elemental and Surface Analysis

ICP measurements were acquired on a CFx and Pb electrode after discharge to quantify the Pb content in each electrode after discharge. Each electrode was placed into deionized water, where soluble Pb ions from the lead fluoride discharge products were hydrolyzed and quantified. The CFx and Pb electrodes were found to contain hydrolyzed PbF2. As discussed above, we estimate that 5.4% of the discharge capacity was due to Pb2+ ion transport, with the balance of discharge capacity (94.6%) proceeding by F– ion transport.

ICP acquired on a discharged Sn electrode reveals that the Sn fluoride discharge products contain 4.77 mg of Sn and 1.52 mg of fluorine. The estimated mass of fluorine removed from CFx based on the specific capacity achieved is 1.58 mg. ICP was not acquired on the CFx electrode; however, the good agreement between the hydrolyzed metallic fluoride content in the Sn anode and electrochemical measurements indicates that at least 96.2% of the discharge product was formed in the Sn anode.

XRF spectroscopy was used to determine and compare the elemental composition of a pristine CFx electrode and a CFx electrode discharged to 0 V. Results reveal a nondetectable weight percentage of Sn metal in the pristine CFx cathode and 2.5 wt % once discharged. The fluorine content of the pristine CFx electrode was 37.2 wt %. Once discharged, the fluorine content decreased to 6.2 wt %, indicating an 83.3 wt % decrease in fluorine content after discharge of the CFx electrode. Given the composition of the electrode, the fluorine content solely from the PVDF binder can be estimated to be 1.69 wt %. After accounting for the binder, the estimated mass of fluorine removed from the sample based on electrochemical measurements is 3.10 mg. From the XRF analysis, the estimated fluorine content removed was 3.51 mg, a value in good agreement with our electrochemical results. Solid-state NMR measurements of the harvested Sn electrode from the Sn-CFx cell revealed that SnF4 was present at the CFx electrode. We estimate that 3% of the discharge capacity was achieved by Sn2+ transport. Estimates of the quantity of metal fluorides present on the CFx electrodes through ICP, XRF, and solid-state NMR are further described in Text S2.

Raman scattering from pristine CFx (Figure 5A) shows a strong D peak at 1360 cm–1, G′ peak at 1595 cm–1, and D + G peak at 2930 cm–1 with strong luminescence from the samples. Raman features of carbon and its derivatives have been extensively studied, and the D peak defines the presence of defects, e.g., lattice dislocation, vacancies, etc., for its activation. The D peak usually appears on the high-frequency shoulder of the G peak, and it activates upon a symmetry break in the carbon plane.

Figure 5 Raman scattering from (A) pristine CFx electrode and (B) discharged CFx electrode. Raman scattering from a (C) discharged Sn electrode was performed at three different positions. Both the CFx and Sn discharged electrodes were harvested from an F-ion Sn-CFx cell.

We also performed Raman scattering measurements on a harvested CFx cathode discharged against a Sn counter electrode (Figure 5B). In the case of discharged CFx, the luminescence is considerably quenched, and D, G, and D + G peaks broaden and are clearly visible in co-occurrence with a significant increase of the D/G peak ratios in comparison to pristine CFx. The D peak arises only from clusters of sp2 sites, and the G peak arises from vibrations of all sp2 sites in both chain and ring configurations. However, some more peaks arise at 256, 554, and 727 cm–1, showing the changes in Raman spectra are most likely caused by the structural disordering in the carbon lattice, e.g., the transformation of sp2 carbon to sp3 hybridization because of the fluoride-ion adsorption. The Raman signature of CFx after discharge exhibits a characteristic similar to that from highly disordered or nanostructured carbon-based materials.

We also investigated the Raman spectra of the discharged Sn electrode at three positions of the electrode surface (Figure 5C). We did not find any characteristic peak of SnO2 at any point. As shown in Raman spectra, we found SnF2 peaks at 387 cm–1 (A1g), 124 cm–1 (Eg), and 82 cm–1 (B1g) due to crystallized SnF2 (Figure 5C, second position). At some other points, we have found a strong peak at 335 cm–1, which is believed to be due to the amorphous structure of SnF2, suggesting formations of nano-ordered regions (Figure 5C, third position). The Raman modes at lower frequencies are due to the amorphous nature of the material.

Molecular-Level Compositions and Environments

Solid-state 19F spin-echo MAS and 119Sn{19F} CP-MAS NMR experiments were acquired on pristine and discharged Sn and CFx electrodes to characterize the molecular-level environments of the fluorine and tin discharge products. To aid signal assignments, commercially purchased SnF2 and SnF4, as well as PVDF, were also characterized.

The solid-state 19F NMR spectrum of pristine PVDF powder (Figure 6A) exhibits five 19F signals at −82, −91, −96, −113, and −115 ppm, revealing multiple characteristic environments associated with different local PVDF structures.20 The dominant 19F resonance at −91 ppm is assigned to amorphous PVDF domains. The 19F signals at −82 and −96 ppm are associated with crystalline PVDF environments, while those at −113 and −115 ppm are assigned to regio-irregular structures.20 The solid-state 19F NMR spectrum of the composite pristine Sn electrode (Figure 6B), which contains 10 wt % PVDF binder, exhibits similar 19F signals to PVDF. However, the 19F signals broadened significantly, a consequence of increased structural disorder due to the ball milling process that was used to prepare the electrode. A weak 19F signal was also observed at −128 ppm, likely associated with an impurity formed during the ball milling process.

Figure 6 Solid-state 19F spin-echo MAS NMR spectra of (A) PVDF, (B) pristine Sn electrode, (C) discharged Sn electrode from the F-ion Sn-CFx cell, (D) commercial SnF4, and (E) commercial SnF2. Spinning sidebands are labeled with asterisks. The SnF4-based impurity in panel (D) is marked with a dagger. All spectra were acquired at 14.1 T and 40 kHz MAS, except for SnF4, which was acquired at 38 kHz MAS.

After electrochemical discharge to 0 V, the solid-state 19F NMR spectrum of the discharged Sn electrode (Figure 6C) reveals a dominant, well-resolved 19F signal at −145 ppm. As shown below, this 19F signal is associated with SnF4. A weaker 19F signal centered at −52 ppm is also observed, which is likely that of SnF2. The discharged composite electrodes were rinsed with ethanol prior to performing solid-state NMR measurements to eliminate any mobile species from the residual electrolyte salt, Np1F. The solid-state 19F NMR spectra of the rinsed and unrinsed electrodes are nearly identical (Figure S2), except for the presence of the Np1F, which was successfully removed by ethanol rinsing.

The solid-state 19F NMR spectra of commercially purchased SnF4 and SnF2 fluorides were acquired to identify the fluorine environments in the discharged Sn electrode. The solid-state 19F NMR spectrum of SnF4 (Figure 6D) reveals an intense, well-resolved 19F signal at −149 ppm, a broad signal centered at −157 ppm, and multiple weaker 19F signals between −120 and −175 ppm. The SnF4 structure has two crystallographic fluorine sites corresponding to bridging and terminal fluorine atoms; thus, only two 19F NMR signals are expected. Notably, this solid-state 19F NMR spectrum is similar to the one shown in Dorn et al., who also reported multiple 19F sites in this spectral region.21 A solid-state 2D 19F{19F} dipolar-correlation NMR experiment (Figure S3) reveals that these 19F signals are all within closer molecular proximity and thus are within the same phase.21 Clearly, significant structural and chemical disorder is present within SnF4, which could be the result of different polymorphic forms or impurities incorporated during the synthesis or handling.

The solid-state 19F NMR spectrum of SnF2 (Figure 6E) shows well-resolved 19F signals at −29, −34, −39, and −45 ppm as well as a partially resolved 19F signal at −41 ppm, which are associated with SnF2. In addition, a prominent 19F signal is observed at −114 ppm, ascribed to SnF4-based impurities. The 19F signals in the region of −30 to −50 ppm are indeed SnF2, as shown by Dorn et al. using 2D solid-state 19F{119Sn} through-bond correlation NMR experiments. Bräuniger et al. and van Wüllen et al. also resolved SnF219F environments, respectively, in the same 19F chemical shift region.22,23 The SnF2 structure has four crystallographic fluorine sites;23 the appearance of at least one additional 19F signal in this region indicates some degree of local disorder.

Interestingly, Dorn et al. characterized SnF2 from two different chemical suppliers: one exhibited the expected 19F signals from SnF2, and the other—which had a solid-state 19F NMR spectrum very similar to the one here (Figure 6E)—had a dominant 19F signal near −110 ppm, which was assigned to SnF4-based impurities due to the 19F chemical shift and 119Sn chemical shift anisotropies.21 Here, the 19F signal at −114 ppm is also suspected to be derived from SnF4-based impurities. A solid-state 2D 19F{19F} NMR dipolar-correlation experiment (Figure S4) reveals 19F signals of SnF2 and impurities within the same phase, in agreement with local disorder. Additionally, a solid-state 119Sn{19F} HETCOR NMR spectrum (Figure S6) reveals a correlation between the 19F signal at −114 ppm and the 119Sn signals centered at −820 ppm.

Solid-state 119Sn{19F} CP-MAS NMR measurements of the discharged Sn electrode were acquired to characterize 119Sn environments in subnanometer-scale proximity to 19F environments. The solid-state 119Sn{19F} CP-MAS NMR spectrum of the discharged Sn anode (Figure 7A) reveals a single 119Sn signal at −803 ppm. The 119Sn chemical shift is distinct compared to commercial SnF4 and SnF2 but notably is the same chemical shift region. The differences in the 119Sn chemical shift among these compounds are due to small changes in the local environment of Sn. A solid-state 2D 19F{19F} dipolar-correlation NMR experiment (Figure S5) also confirms that significant structural disorder is present within the fluorinated tin discharge products.

Figure 7 Solid-state 119Sn{19F} CP-MAS NMR of a (A) discharged Sn anode F-ion Sn-CFx cell, (B) commercial SnF4, and (C) commercial SnF2. In panel (B), spinning sidebands associated with the 119Sn SnF4 signal at −767 ppm are labeled with black asterisks. In panel (C), spinning sidebands for SnF2 sites 1 and 2 are labeled with black and red asterisks, respectively. Impurities are marked with daggers. All spectra were acquired at 14.1 T and 40, 38, and 30 kHz MAS, respectively.

The solid-state 119Sn{19F} CP-MAS NMR spectrum of SnF4 (Figure 7B) reveals a dominant 119Sn signal at −767 ppm with a broad 119Sn shoulder at approximately −757 ppm, a weak, broad 119Sn signal at −785 ppm, and a sharp 119Sn signal at −858 ppm. The solid-state 119Sn{19F} CP-MAS spectrum is very similar to the one reported by Dorn et al. and Bräuniger et al. SnF4 contains two crystallographic tin sites; the 119Sn signals at −757 and −785 ppm are assigned to tin environments in coordination with terminal (site 1) and bridging (site 2) fluorine sites, respectively.21 Here, the 119Sn signals at −757 and −767 ppm may both be associated with terminal sites in slightly different local environments. The sharp 119Sn signal at −858 ppm is likely due to an impurity in a well-defined local environment.

The solid-state 119Sn{19F} CP-MAS NMR spectrum of commercial SnF2 (Figure 7C) reveals two isotropic 119Sn signals at −934 and −1039 ppm assigned to tin environments coordinated with three (site 1) or five fluorine (site 2) sites. SnF2 contains 119Sn in local coordination environments with either three or five fluorine atoms; thus, two isotropic 119Sn signals are expected.21 The intense, well-defined 119Sn signal at −820 ppm is ascribed to SnF4-based impurities, as discussed above, and is associated with the solid-state 19F spin-echo NMR spectrum signal at −114 ppm.

Solid-state 2D dipolar mediated techniques were used to selectively probe interactions between 19F signals that are dipole–dipole coupled and, thus, in close molecular proximity to the 119Sn signals of the discharged Sn electrode. A solid-state 119Sn{19F} HETCOR NMR spectrum of the discharged Sn electrode (Figure 8) reveals a 2D correlated signal between the 119Sn at −803 ppm and 19F at −145 ppm, providing molecular-level evidence that SnF4 is electrochemically formed upon discharge. 2D correlations were not observed between any 119Sn signals and the 19F signal at −52 ppm associated with SnF2; this is likely the result of the lower concentration of the SnF2 species in the bulk Sn electrode.

Figure 8 Solid-state NMR experiments performed on a discharged Sn electrode from a F-ion Sn-CFx cell acquired at 40 kHz MAS and 14.1 T. (A) 2D 119Sn{19F} HETCOR spectrum, (B) separately acquired 119Sn{19F} CP-MAS spectrum, (C) 19F internal projection of the 2D HETCOR spectrum, and (D) separately acquired 19F spin-echo NMR spectrum.

The solid-state 19F NMR spectrum of a pristine CFx electrode (Figure 9A) reveals two intense 19F signals at −184 and −170 ppm attributed to covalent and semi-ionic CF bonds.24 Semi-ionic C–F bonds may also be described as covalent in nature but with a contribution from hyperconjugation.25,26 The C–F bond order is lowered within the CFx structure due to hyperconjugation within the C–C bonds, resulting in a distortion from its covalent nature.27 This distortion changes the local fluorine electronic environment, altering the 19F NMR chemical shift. The broad 19F signal centered at −112 ppm and multiple well-resolved low-intensity signals between −70 and −90 ppm are associated with CF2 and CF3 sites. The 19F signal centered at −91 ppm corresponds to the amorphous domain of PVDF. As previously discussed, the PVDF binder exhibits multiple 19F signals depending on the local structure. The CFx composite electrode consists of only 3 wt % PVDF; thus, the PVDF signals are weak in intensity.

Figure 9 Solid-state 19F spin-echo MAS NMR spectra of (A) CFx electrode and (B) discharged CFx electrode from a F-ion Sn-CFx cell. Spinning sidebands are labeled with an asterisk. All spectra were acquired at 14.1 T and 40 kHz MAS.

Upon discharge, the 19F NMR spectrum of the discharged CFx electrode (Figure 9B) reveals that 19F signals of CF covalent and semi-ionic fluorine environments decreased significantly with respect to the pristine CFx electrode, as observed by the decrease in their relative 19F signal intensities. Critically, these results establish at the molecular level that CFx is electrochemically defluorinated upon discharge. A well-resolved 19F signal at −145 ppm associated with SnF4 is observed, establishing that the mass transport of tin can occur from the Sn anode to the CFx cathode. SnF2 was not observed in the discharged CFx electrode, likely due to its low concentration in the bulk electrode sample. Sn transport was detected by XRF and solid-state NMR but not by XRD or Raman measurements. This apparent discrepancy may be due to the nature of the different measurement techniques and the amorphous nature of the electrochemically formed SnF4: XRD detects only reflections of crystalline components, while Raman scattering is limited to a specific region and depth resolution on the electrode surface.

To enable the relative 19F populations to be quantified, the solid-state 19F MAS NMR spectra of pristine CFx and discharged CFx and Sn electrodes were deconvoluted into individual 19F signals. The discharged Sn and CFx electrodes were harvested from a cell that achieved a specific capacity of 314.9 mAh g–1, where the fraction of capacity extracted from the cell relative to the theoretical value of CFx (865 mAh g–1, for x = 1) is 36.4%. The 19F molar populations of covalent and semi-ionic CF environments relative to the PVDF binder were 83 and 44% for the pristine and discharged CFx electrodes, respectively, or a decrease of 39% (Figure 9). Note that the full theoretical specific capacity was not achieved by using these in-house Sn-CFx cells, which were not experimentally optimized to extract maximum capacity. The 19F molar percentage decrease of 39% in the CFx is in excellent agreement with the fraction of capacity extracted from the cell. Moreover, the ca. 2.6% discrepancy between the extracted capacity and the decrease in the 19F CF molar percentage in the CFx electrode confirms that ∼3% of the discharge capacity was achieved by Sn2+ transport. The 19F molar population percentages of SnF2 and SnF4 present in the discharged Sn electrode (8.6 and 91.4%, respectively) were compared to the percentage of capacity extracted during the two plateaus observed during galvanostatic discharge (8.4 and 91.6%, respectively). The first and second discharge plateaus are therefore ascribed to the electrochemical formation of SnF2 and SnF4 from Sn metal, respectively, as they are in agreement with the quantified 19F molar populations.

The electrochemical discharge of an F-ion metal-CFx cell is summarized schematically in Figure 10. The conversion mechanism of an F-ion metal anode-CFx cell is the cathodic reduction of the CFx electrode to form fluoride ions (F–) and carbon (C) (eq 1)1

the transport of fluoride ions through the electrolyte, and the anodic oxidation of the metal electrode (M) to form metal fluorides (MF) with the discharge products remaining at the anode (eq 2).2

The overall electrochemical reaction is shown in eq 3.3

Figure 10 Schematic illustrating the electrochemical discharge of an F-ion metal-CFx primary (nonrechargeable) cell.

Conclusions

In this work, we demonstrate the electrochemical defluorination of CFx cathodes using a room-temperature F-ion conducting electrolyte when paired with either a Pb or Sn anode. The primary F-ion Pb-CFx and Sn-CFx cells achieved capacities of 700 and 400 mAh g–1, respectively, without optimization. XRD measurements show that the metal fluorides PbF2 or SnF2 form on the Pb or Sn anodes, respectively. Solid-state 19F and 119Sn{19F} NMR measurements of an F-ion Sn-CFx discharged Sn electrode and electrode components revealed the presence of both SnF2 and SnF4, where SnF4 is the main discharge product. Solid-state 2D 19F{19F} dipolar-correlation NMR experiments of the discharged Sn electrode revealed the amorphous nature of the electrochemically formed SnF4, confirming why it was not observed by XRD. The presence of nano-ordered regions, with some amorphous in nature, is also validated by Raman scattering and surface morphology changes of the electrodes upon discharge.

Metal fluorides were discovered on the discharged CFx cathode of both the F-ion Pb-CFx and Sn-CFx cells, as evidenced by XRD, XRF, and solid-state NMR. ICP, XRF, and solid-state NMR quantified a ca. 5.4 and 3.1% contribution to the discharge capacity due to Pb2+ and Sn2+ transport and subsequent reaction. Thus, a >40% reduction in metal fluoride transport to the CFx electrode was achieved when using Sn, compared to the Pb. Solid-state 19F NMR measurements of a pristine and discharged CFx electrode from a Sn-CFx cell establish unambiguously that the C–F bonds are broken upon discharge. The 19F molar percentage of CF bonds consumed (39%) was compared to the cell capacity extracted related to the theoretic capacity of CFx (36.4%), confirming the electrochemical defluorination of CFx as well as a minor contribution due to Sn cation transport. Overall, we have demonstrated a new electrochemical discharge mechanism relative to conventional Li-CFx cells that reduces polarization loss and raises the possibility of improvement to a higher practical discharge voltage. Future work will address the lithium metal counter electrode reactivity to enable an F-ion-based Li-CFx cell.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.jpcc.4c03412.Structural parameters obtained from the Rietveld refinements, XRD patterns of electrode components and electrodes, solid-state 19F NMR spectra of electrolyte salt and electrodes, solid-state 2D 19F{19F} and119Sn{19F} MAS experiments, supporting text, and calculations (PDF)

Supplementary Material

jp4c03412_si_001.pdf

The authors declare no competing financial interest.

Acknowledgments

The authors gratefully acknowledge funding from the U.S. National Aeronautics and Space Administration (NASA) via the NASA-CCNY Center for Advanced Batteries for Space under cooperative agreement #80NSSC19M0199. Part of this research was carried out at the Jet Propulsion Laboratory, California Institute of Technology, under a contract with NASA (#80NM0018D0004). This work was supported in part by a grant from Flion Energy Inc. J.L.A. acknowledges support from the U.S. National Science Foundation (NSF) Graduate Research Fellowship under grant No. DGE-1842487. The authors thank Bill Warner from the Jet Propulsion Laboratory (JPL) for his assistance with ICP and XRF measurements. Solid-state NMR measurements were performed at the City University of New York (CUNY) Advanced Science Research Center. Raman scattering measurements were performed at the Institute for Functional Nanomaterials of the University of Puerto Rico with support from NASA under cooperative agreement #80NSSC19M0236. B.C.M. receives research funding from Flion Energy Inc., which is developing technologies related to the research described in this paper. In addition, the author serves as a consultant to Flion Energy Inc. and holds equity in the company in compensation for these services. B.C.M.’s relationship with Flion Energy Inc. has been reviewed and approved by the University of Southern California in accordance with its conflict-of-interest policies.
==== Refs
References

Read J. ; Collins E. ; Piekarski B. ; Zhang S. LiF Formation and Cathode Swelling in the Li/CF x Battery. J. Electrochem. Soc. 2011, 158 (5 ), A504–A510. 10.1149/1.3561429.
Linden D. ; Reddy T. B. In Handbook of Batteries, 3rd ed.; Thomas R. , Ed.; McGraw-Hill, 2001.
Jones J. P. ; Jones S. C. ; Billings K. J. ; et al. Radiation effects on lithium CFX batteries for future spacecraft and landers. J. Power Sources 2020, 471 , 228464 10.1016/j.jpowsour.2020.228464.
Smith P. H. ; Sepe R. B. ; Waterman K. G. ; Myron L. J. Development and Analysis of a Lithium Carbon Monofluoride Battery-Lithium Ion Capacitor Hybrid System for High Pulse-Power Applications. J. Power Sources 2016, 327 , 495–506. 10.1016/j.jpowsour.2016.07.035.
Hand K. P. ; Phillips C. B. ; Murray A. ; et al. Science Goals and Mission Architecture of the Europa Lander Mission Concept. Planet. Sci. J. 2022, 3 (1 ), 22 10.3847/PSJ/ac4493.
Krause F. C. ; Jones J. P. ; Jones S. C. ; et al. High Specific Energy Lithium Primary Batteries as Power Sources for Deep Space Exploration. J. Electrochem. Soc. 2018, 165 (10 ), A2312–A2320. 10.1149/2.1061810jes.
Krishnamurthy V. ; Viswanathan V. Beyond Transition Metal Oxide Cathodes for Electric Aviation: The Case of Rechargeable CFx. ACS Energy Lett. 2020, 5 (11 ), 3330–3335. 10.1021/acsenergylett.0c01815.
Leung K. ; Schorr N. B. ; Mayer M. ; Lambert T. N. ; Meng Y. S. ; Harrison K. L. Edge-Propagation Discharge Mechanism in CFx Batteries - A First-Principles and Experimental Study. Chem. Mater. 2021, 33 (5 ), 1760–1770. 10.1021/acs.chemmater.0c04676.
Whittingham M. S. Mechanism of Reduction of the Fluorographite Cathode. J. Electrochem. Soc. 1975, 122 (4 ), 526–527. 10.1149/1.2134252.
Touhara H. ; Fujimoto H. ; Watanabe N. ; Tressaud A. Discharge Reaction Mechanism in Graphite Gluoride-Lithium Batteries. Solid State Ionics 1984, 14 (2 ), 163–170. 10.1016/0167-2738(84)90093-6.
Guérin K. ; Dubois M. ; Hamwi A. Advanced Carbon Fluorides for Primary Lithium Batteries. E3S Web Conf. 2017, 16 (1 ), 17002 10.1051/e3sconf/20171617002.
Watanabe N. Two Types of Graphite Fluorides, (CF)n and (C2F)n, and Discharge Characteristics and Mechanisms of Electrodes of (CF)n and (C2F)n in Lithium Batteries. Solid State Ionics 1980, 1 (1–2 ), 87–110. 10.1016/0167-2738(80)90025-9.
Thieu D. T. ; Fawey M. H. ; Bhatia H. ; et al. CuF2 as Reversible Cathode for Fluoride Ion Batteries. Adv. Funct. Mater. 2017, 27 (31 ), 1701051 10.1002/adfm.201701051.
Konishi H. ; Minato T. ; Abe T. ; Ogumi Z. Reversible Electrochemical Reaction of a Fluoride Shuttle Battery with a Bismuth(III) Fluoride Electrode and Electrolyte Containing Triphenylboroxine as an Anion Acceptor. ChemistrySelect 2020, 5 (21 ), 6237–6241. 10.1002/slct.202001163.
Davis V. K. ; Bates C. M. ; Omichi K. ; et al. Room-temperature cycling of metal fluoride electrodes: Liquid electrolytes for high-energy fluoride ion cells. Science 2018, 362 (6419 ), 1144–1148. 10.1126/science.aat7070.30523107
Davis V. K. ; Munoz S. ; Kim J. ; et al. Fluoride-ion solvation in non-aqueous electrolyte solutions. Mater. Chem. Front. 2019, 3 (12 ), 2721–2727. 10.1039/C9QM00512A.
Coelho A. A. TOPAS and TOPAS-Academic: An optimization program integrating computer algebra and crystallographic objects written in C++. J. Appl. Crystallogr. 2018, 51 (1 ), 210–218. 10.1107/S1600576718000183.
Massiot D. ; Fayon F. ; Capron M. ; et al. Modelling One- and Two-Dimensional Solid-State NMR Spectra. Magn. Reson. Chem. 2002, 40 (1 ), 70–76. 10.1002/mrc.984.
Linden D. ; Reddy T. B. Handbook of Batteries; McGraw-Hill, 1995; Vol. 33 .
Koseki Y. ; Aimi K. ; Ando S. Crystalline structure and molecular mobility of PVDF chains in PVDF/PMMA blend films analyzed by solid-state 19F MAS NMR spectroscopy. Polym. J. 2012, 44 (8 ), 757–763. 10.1038/pj.2012.76.
Dorn R. W. ; Carnahan S. L. ; Cheng C.-y. ; Pan L. ; Hao Z. ; Rossini A. J. Structural characterization of tin in toothpaste by dynamic nuclear polarization enhanced 119Sn solid-state NMR spectroscopy. Nat. Commun. 2023, 14 (1 ), 7423 10.1038/s41467-023-42816-z.37973961
Bräuniger T. ; Ghedia S. ; Jansen M. Covalent bonds in α-SnF2 monitored by J-couplings in solid-state NMR spectra. Z. Anorg. Allg. Chem. 2010, 636 (13–14 ), 2399–2404. 10.1002/zaac.201000176.
van Wüllen L. ; Schiffmann J. G. ; Kopp J. ; et al. Development and application of novel NMR methodologies for the in situ characterization of crystallization processes of metastable crystalline materials. Z. Kristallogr.—Cryst. Mater. 2017, 232 (1–3 ), 141–159. 10.1515/zkri-2016-1975.
Schoetz T. ; Robinson L. E. ; Gordon L. W. ; et al. Elucidating the Role of Electrochemically Formed LiF in Discharge and Aging of Li-CFx Batteries. ACS Appl. Mater. Interfaces 2024, 16 (15 ), 18722–18733. 10.1021/acsami.3c17562.38587415
Kim K. H. ; Cho J. H. ; Hwang J. U. ; Im J. S. ; Lee Y. S. A Key Strategy to Form a LiF-based SEI Layer for a Lithium-Ion Battery Anode With eEhanced Cycling Stability by Introducing a Semi-Ionic C–F Bond. J. Ind. Eng. Chem. 2021, 99 , 48–54. 10.1016/j.jiec.2021.04.002.
Chen P. ; Liu W. ; Wang H. ; Jiang Y. ; Niu X. ; Wang L. Semi-Ionic C-F Bond Enabling Fluorinated Carbons Rechargeable as Li-Ion Batteries Cathodes. J. Colloid Interface Sci. 2023, 649 , 255–263. 10.1016/j.jcis.2023.06.108.37348345
Giraudet J. ; Dubois M. ; Guérin K. ; Delabarre C. ; Hamwi A. ; Masin F. Solid-state NMR Study of the Post-Fluorination of (C2.5F)n Fluorine - GIC. J. Phys. Chem. B 2007, 111 (51 ), 14143–14151. 10.1021/jp076170g.18052273
