- Email: [email protected]
Designer Anion Enabling Solid-State Lithium-Sulfur Batteries
Article
Designer Anion Enabling Solid-State LithiumSulfur Batteries Heng Zhang, Uxue Oteo, Xabier Judez, ..., Javier Carrasco, Chunmei Li, Michel Armand [email protected] (H.Z.) [email protected] (M.A.)
HIGHLIGHTS Scalable designer anion is proposed for solid-state Li-S batteries (SSLSBs) New designer anion shows superior stability against Li metal electrode Long-term cycling of SSLSBs is achieved with designer anion Structural modification of salt anions tunes the performance of Li metal batteries
In this work, we developed a new designer anion favoring Li-ion transport and improving interfacial compatibility with Li metal electrode, thus boosting the performance of solid-state Li-S batteries (SSLSBs). The designer anion could be made with the industrially available intermediates and is believed to expedite the development of SSLSBs as well as other rechargeable Li batteries.
Zhang et al., Joule 3, 1689–1702 July 17, 2019 ª 2019 Elsevier Inc. https://doi.org/10.1016/j.joule.2019.05.003
Article
Designer Anion Enabling Solid-State Lithium-Sulfur Batteries Heng Zhang,1,* Uxue Oteo,1 Xabier Judez,1 Gebrekidan Gebresilassie Eshetu,2,3 Maria Martinez-Iban˜ez,1 Javier Carrasco,1 Chunmei Li,1 and Michel Armand1,4,*
SUMMARY
Context & Scale
With an extremely high theoretical energy density, solid-state lithium-sulfur (Li-S) batteries (SSLSBs) are emerging as one of the most feasible chemistries; however, their energy efficiency and long-term cyclability are severely hampered by the lithium metal (Li ) dendrite formation during repeated discharge/charge cycles and the shuttling of aggressive polysulfide intermediates between two electrodes. Herein, we report (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide anion [N(SO2CF2H)(SO2CF3)], hereafter DFTFSI, as a designer anion for high-performance polymer-based SSLSBs. In contrast to the widely used bis(trifluoromethanesulfonyl)imide anion [N(SO2CF3)2] (TFSI), DFTFSI-based SSLSBs provide superior interfacial stability against Li , extremely high discharge and areal capacities, very high Coulombic efficiency, and long-term cyclability, surpassing the reported literature values, in terms of gravimetric energy density. This work opens a new door for accelerating the practical deployment of SSLSBs in the future.
Solid-polymer-electrolyte (SPE)based solid-state Li-S batteries (SSLSBs) are endowed with profuse beneficial futures such as high theoretical energy density, low cost, higher safety, and environmental benignancy. However, their practical deployment is hindered by several intrinsic problems, e.g., Li metal (Li ) dendrite or mossy growth, electronically insulating nature of S8 and Li2S, polysulfides shuttling, etc. This work presents a breakthrough in improving the performance of SSLSBs via a designer anion, (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide anion (DFTFSI). DFTFSI-based electrolytes show dendrite-free Li plating and stripping, and enable the long-term cycling of LiS cells with high capacities and excellent Coulombic efficiencies. Thus, this study opens a new avenue toward the design of new and tailored SPEs for applications in a high-performance and safer Li-S battery as well as other rechargeable Li batteries.
INTRODUCTION Sustainable development of today’s society puts forward an increasing demand for efficient and clean usage of energy sources.1,2 Lithium (Li)-ion batteries (LIBs), an important part for building the electrified network, have been widely deployed as power sources for smart electronic and (hybrid) electric vehicles (xEVs), as well as buffers for intermittent renewable energy (e.g., wind, solar).3–7 With a ceiling gravimetric energy of ca. 400 Wh kg1, LIBs still fall behind the requirement of future xEVs and other emerging applications.8–12 Among all the post Li-ion chemistries, Li-sulfur (Li-S) batteries appear to be the most appealing and viable energy storage technologies in view of their extraordinary energy density (theoretical value of ca. 2,600 Wh kg1), abundant sulfur resources, cost effectiveness, and environmental benignancy.8,13,14 However, the practical implementation of Li-S batteries faces formidable challenges, such as the low electronic conductivity of elemental sulfur (S8) and Li sulfide (Li2S), shuttling effect of the reaction intermediates (i.e., polysulfide [PS]), and the notorious Li metal (Li ) dendrite formation bearing along low Coulombic efficiencies and safety concerns in the presence of volatile liquid ethers solution.8,13–18 Solid polymer electrolytes (SPEs), first found by Wright and coworkers19 in 1973 and then suggested by Armand20 as electrolytes for rechargeable batteries in 1978, have been recognized as promising candidates for accessing safe and high-performance rechargeable Li -based batteries.21–24 This is motivated by their intrinsically enhanced safety compared to conventional liquid electrolytes, good processability, and flexibility compared to inorganic solid electrolytes.21–24 Besides, the
Joule 3, 1689–1702, July 17, 2019 ª 2019 Elsevier Inc. 1689
replacement of liquid electrolytes with SPEs could remarkably enhance the energy density of Li-S batteries as shown in recent calculations.25,26 As one of the most critical parameters, the nature of salt anions plays a pivotal role in dictating the electrochemical performance of solid-state Li-S batteries (SSLSBs).26–28 To date, bis(trifluoromethanesulfonyl)imide anion {[N(SO2CF3)2], TFSI} has captured the most interest among all the anions evaluated for SSLSBs because of its good thermal and chemical stability, structural flexibility, and plasticizing effect when coupled with the widely used poly(ethylene oxide) (PEO) matrix.21–24 However, the solid-electrolyte interphase (SEI) layers formed between the LiTFSI/PEO electrolyte and Li anode are not stable enough toward the aggressive PS intermediates, thereby resulting in an inferior cycling performance of SSLSBs.29–32 The (fluorosulfonyl)(trifluoromethanesulfonyl)imide anions {[N(SO2F)(SO2CF3)], FTFSI} revealed in our recent work clearly shows the importance of designer anion on the electrochemical performance of SSLSBs.32 The corresponding Li-S cells delivered high specific and areal capacity with good Coulombic efficiency (CE) because of the synergistic effects resulting from its molecular structure (i.e., possessing both SO2CF3 and SO2F functionalities). However, LiFTFSI is susceptible to water, aprotic solvents, and heat, limiting the scalable preparation of its electrolytes, and the stability of the Li electrode still needs to be improved for long-term cycling of SSLSBs. In this work, we demonstrate that a designer anion, (difluoromethanesulfonyl)(trifluoromethanesulfonyl)imide {[N(SO2CF2H)(SO2CF3)], DFTFSI}, hoists the performance of polymer-based SSLSBs. The chemical structure of DFTFSI is designed on the basis of the following considerations (schematically illustrated in Figure 1). Firstly, the CF2H moiety, being weakly acidic, is beneficial for restricting anionic mobility via the hydrogen-bonding interaction between DFTFSI– and PEO, resulting in an enhanced Li-ion conductivity (sLi+ = 2 3 104 S cm1 [LiDFTFSI/PEO] versus sLi+ = 1.5 3 104 S cm1 [LiTFSI/PEO] at 70 C33,34). Of note is that higher anionic mobility is responsible for the steeper concentration gradient accumulated during the discharge and charge processes, which causes an increased polarization, voltage loss, and cell failure of SSLSBs.15,24,35 Secondly, the asymmetric structure of DFTFSI– and the presence of electrochemically labile CF2H moiety might induce a facile decomposition of DFTFSI– on a Li anode, leading to the formation of several important SEI building species, such as ionic conductive Li hydride (LiH [s = 10–10 S cm–1]36 versus Li2S [s = 10–13 S cm–1]37 at 25 C) and mechanically stable Li fluoride. Lastly, the preparation of DFTFSI– could be made in a sequence of high-yield reactions with the industrially available intermediates (e.g., difluoromethanesulfonyl chloride from Solvay), thus ensuring the potential large-scale application of LiDFTFSI in SSLSBs.
RESULTS The Chemistry of Designer Anion with Li Electrode Firstly, the effect of salt concentration on the phase-transition behavior and ionic conductivity of LiDFTFSI/PEO is investigated prior to the understanding of anion chemistry on Li electrode. As seen in Figure S1, at the mole ratio of ethylene oxide (EO)/Li = 20, LiDFTFSI/PEO shows a low glass transition temperature (Tg) at 38 C. With the addition of Li salt, the value of Tg increases gradually up to 30 C for the sample of EO/Li = 8. As a consequence, the ionic conductivity of LiDFTFSI/PEO shows the highest at EO/Li = 20 and drops at a higher salt content (see Figure S2). Therefore, LiDFTFSI- and LiTFSI-based electrolytes at EO/Li = 20 were selected for the following electrochemical studies. In addition, as displayed in Figure S3, the
1690 Joule 3, 1689–1702, July 17, 2019
1Electrical
Energy Storage Department, CIC Energigune, Parque Tecnolo´gico de A´lava, Albert Einstein 48, 01510 Min˜ano, A´lava, Spain 2Institute
for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Ja¨gerstraße 17/19, 52066 Aachen, Germany
3Department
of Chemistry, College of Natural and Computational Sciences, Mekelle University, P.O. Box 231, Mekelle, Ethiopia
4Lead
Contact
*Correspondence: [email protected] (H.Z.), [email protected] (M.A.) https://doi.org/10.1016/j.joule.2019.05.003
Figure 1. The Role of LiDFTFSI and LiTFSI in SSLSBs
anodic stability of LiDFTFSI/PEO and LiTFSI/PEO is higher than 4 V versus Li/Li+, which is well acceptable for the operation of Li-S batteries. Highly reversible and dendrite-free plating and stripping of Li electrode is crucial for the long-term cyclability, rate capability, CE, and safety of SSLSBs. The electrochemical stabilities of Li electrode in both LiDFTFSI- and LiTFSI-based electrolytes are comparatively displayed in Figure 2. As shown in Figure 2A, the galvanostatic cycling of Li symmetric cells using the LiDFTFSI/PEO electrolyte sustains more than 600 h without any erratic voltage oscillation, whereas the LiTFSI-based one encounters an internal short circuit after 168 h of cycling.38 The excellent stability of Li electrode in the LiDFTFSI-based electrolyte is further manifested by the superior cycling stability of Li symmetric cells using the corresponding liquid electrolyte, where the cycle life of Li j LiDFTFSI/1,2-dimethoxyethane (DME) j Li cell is 25 times higher than that of the LiTFSI-based one (Figure 2B, > 1,000 h [LiDFTFSI/DME] versus ca. 40 h [LiTFSI/DME]32). The markedly enhanced cyclability of the LiDFTFSI-based cell strongly evidences the formation of stable SEI layers on Li electrode. In addition, the slightly higher overvoltage in the LiDFTFSI-based cells (both liquid and polymer electrolytes) than in LiTFSI-based ones implies that the SEI layers generated in the former electrolytes tend to be more resistive and compact (see Figures 2D and 2E for X-ray photoelectron spectroscopy [XPS] spectra). To elucidate the role of CF2H moiety on the electrochemical behavior of Li electrode, post-mortem analyses employing scanning electron microscopy (SEM) and XPS were performed in the liquid cell configuration. Note that Li electrode in contact with PEO-based electrolytes cannot be separated without damaging its surface because of the strong adhesive properties of PEO; hence, DME having an analogous chemical structure to PEO was used for assessing the Li j electrolyte interphase.30–32,39 As seen in Figure 2C, the Li deposits on copper substrate (2 mAh cm2 deposition) recovered from the LiDFTFSI-based electrolyte show
Joule 3, 1689–1702, July 17, 2019 1691
Figure 2. Electrochemical Compatibility of Designer Anion with Li Electrode (A and B) Galvanostatic cycling of Li electrode in the PEO-based electrolytes at 70 C (A) and DMEbased electrolytes at 25 C (B). Current density, 0.1 mA cm2 ; half-cycle time, 2 h. (C) Physical appearances and SEM images of Li deposited onto Cu substrates at 0.1 mA cm 2 (plating time 20 h) in LiX/DME (X = DFTFSI or TFSI) electrolytes at 25 C. (D) XPS spectra of the corresponding Li deposits at 25 C. In the figure, R 1 CO and R 2 CO refer to CH3 O and H2 C=HCOCH 3 , respectively; and the black and violet lines correspond to the respective raw and fitted data. (E) Atomic concentration of Li, C, O, F, S, and N on the Li deposits with different etching time using the two liquid electrolytes. For comparison, the raw data of Li plating and stripping and XPS spectra of LiTFSI sample are retreated and presented from Aldalur et al. 38 and Eshetu et al.32 with permission. Copyright 2018 Wiley-VCH Verlag GmbH & Co. KGaA; Copyright 2018 American Chemical Society.
better coverage and homogeneity than the ones from the LiTFSI-based electrolyte, presumably because of the enhanced kinetics for the Li nucleation and growth steps. Besides, smooth, uniform, and pancake-like morphologies of Li deposits are observed for the LiDFTFSI-based electrolyte. In comparison, a large amount of typical needle-like Li dendrites are formed in the LiTFSI-based one.31,32,40 This distinct morphological observation between LiDFTFSI and LiTFSI-based electrolytes clearly suggests that the replacement of CF3 group in TFSI– with a CF2H group
1692 Joule 3, 1689–1702, July 17, 2019
endows the formation of compact and dense Li deposits that are critical for the long-term cycling of SSLSBs. Figure 2D presents the most relevant spectra, C1s and F1s, of the outermost surface layer covered on Li deposits. The assignments of each peak in the XPS spectra as extracted in the previous reports are summarized in Table S1, and the corresponding survey spectra are given in Figure S4. The C1s spectra of both electrolyte systems show dominative peaks in the range of 289 to 284 eV, which are attributed to the presence of alkoxide (e.g., CH3OLi, R1CO) and methoxy-vinyl-ether (e.g., H2C=HCOCH3, R2CO) resulting from the electrochemical reduction of DME on the Li surface. The reduction mechanism of DME has been extensively discussed in our recent work32 as well as previous work from Aurbach et al.41 and Hu et al.42 (Scheme S1), where the scission of the CO bond in DME could form SEI building species such as RCOLi and RCOCR. In addition, the minor signals at slightly higher binding energies (i.e., 294–290 eV) belong to the salt-derived species, such as CF3- and CF2-containing components. Those species are also observed in the corresponding F1s spectra as shown in Figure 2D. Notably, the F1s signal at 685 eV associated with LiF is more prominent for the LiDFTFSI-based electrolyte than that for the LiTFSI-based one (see Table S2 for the normalized percentage of each F-containing species). This suggests that the electrochemical reduction of LiDFTFSI is easier and more exhaustive than that of LiTFSI because of its asymmetric structure and the presence of an electrochemically labile CF2H group. The less negative reduction potential (Ere) for LiDFTFSI is further demonstrated by the chemical simulation experiments with naphthalene radical mono-anions as reducing agents. As shown in Figure S5, LiDFTFSI gets reduced easily upon the addition of naphthalene radical anions (ca. 1 versus Li/Li+43)/tetrahydrofuran (THF) solution, whereas LiTFSI remains stable under the same condition. Previous work revealed that the reduction of LiTFSI is difficult even when treated with biphenyl di-radical anions, which has a potential of ca. 0.15 V versus Li/Li+.31,43 To ascertain the nature and properties of the SEI layer formed in the LiDFTFSI- and LiTFSI-based electrolytes, depth-dependent compositions of several typical elements (Li, C, O, F, S, and N) are shown in Figure 2E. For both electrolytes, the Li concentration increases with increase of the sputtering time with Ar+ ions because of the removal of outermost reduction products (e.g., Li salts and solvents) on the SEI layer. Yet, the Li concentration of the LiDFTFSI-based electrolyte reaches a constant value in a shorter sputtering time (120s [LiDFTFSI] versus 540s [LiTFSI]), suggesting a compact and dense SEI layer formed in LiDFTFSI-based electrolyte. Meanwhile, the F1s spectra gathered from various sputtering times (Figure S6) show that LiF, the ultimate reduction product, becomes a dominant component at the inner part of the formed SEI layer after only 30s sputtering for the LiDFTFSI-based electrolyte. This again proves that the electrochemical reduction of LiDFTFSI promotes the formation of a thin, LiF-rich SEI layer. As shown in Scheme S1, the reduction of LiDFTFSI possibly leads to the formation of ionic conductive SEI building species, LiH, though it could be observed from neither XPS spectra because of the overlapping between LiH and LiOH (unavoidably present as native SEI species on Li electrode) nor Fourier transform infrared spectroscopy (FTIR) spectra as carried out by Aurbach et al.44 because of the concurrent FTIR absorptions of LiH, and SO2 moieties in sulfonimide salts and CO in ethereal solvents. Thus, the combination of mechanically stable LiF and ionic conductive LiH as SEI building materials in the LiDFTFSI-based electrolyte largely inhibits the Li dendrite growth and thereby improves the electrochemical stability of Li electrode.
Joule 3, 1689–1702, July 17, 2019 1693
Figure 3. Mechanistic Understanding of Designer Anion with Li Electrode (A and B) Proposed mechanism for the reduction of the LiDFTFSI (A) and LiTFSI (B) salts. (C) Schematic illustration on the SEI layer formed on Li electrode in both salts.
On the basis of the above-mentioned morphological (Figure 2C) and compositional information (Figures 2D and 2E) of Li electrode surface, together with the chemical reduction simulation tests of the neat Li salts, the SEI layers generated in the LiDFTFSI- and LiTFSI-based electrolytes are schematically depicted in Figure 3. An initial two-electron reduction of DFTFSI– and TFSI– gives alkyl sulfonate or alkyl sulfonylamide intermediates (Figures 3A and 3B), e.g., CF2QSO2Li and CF2QSO2NLi2 (Q = H or F) for DFTFSI–. Sequential reduction of the intermediates affords electrochemically stable inorganic species, such as LiF, LiH, Li2S, and Li3N. Alternatively, these highly reductive species are also accessible with the rigorous degradation of the anions on Li surface, which processes the lowest possible electrochemical potential (–3.040 V versus standard hydrogen electrode [SHE]15). As seen in Figure 3C, the SEI layers covered on Li electrode consist of three parts, including (1) the outermost organic-rich layer emanating from solvent reduction and/or residual salts trapped in the passivation layer, (2) inorganic-rich layer derived from the decomposition of Li salts, and (3) innermost native SEI layer inherited from the Li production and processing. The distinct difference between two electrolytes on the SEI formation lies in a compact and thin inorganic-rich SEI layer formed in the LiDFTFSI-based one, mainly attributed to the facile and exhaustive electro- or chemical reduction of DFTFSI– as a consequence of the replacement of CF3 with the CF2H group in TFSI–, as noted above in Figures 2D, 2E, and S2. The Chemical Stability of the Designer Anion with Polysulfide Since the high-order PS intermediates (Li2Sx, x = 3–8) formed during the discharge and charge process are soluble in PEO-based SPEs, the resistance of salt anions toward those aggressive PS species is of vital importance for the long-term cyclability of SSLSBs.30,32,39 Several Li salts identified in previous works are proved to be unsuited for SSLSBs, despite the notably enhanced stability toward Li anode in other battery systems. For example, Li bis(fluorosulfonyl)imide {Li[N(SO2F)2], LiFSI} was proposed as a potential replacement of LiTFSI in view of its good compatibility
1694 Joule 3, 1689–1702, July 17, 2019
Figure 4. The Chemical Stability of Designer Anion with Polysulfide (A) The appearance of 1 M LiX/DME (X = DFTFSI or TFSI) and blank DME solution before and after the addition of PS at room temperature for 60 h. (B) Normalized UV-vis absorption spectra of the PS-added solutions. (C) DFT calculations for the proposed intermediates. Light gray, red, yellow, light blue, dark blue, gray, and pink balls stand for H, O, S, F, N, C, and Li atoms, respectively.
with Li anode. However, recent works showed that LiFSI reacts spontaneously with PS species, which causes an irreversible consumption of electroactive materials, thus resulting in a decreased discharge capacity of SSLSBs.30,32,39 To this end, the chemical stabilities of both LiDFTFSI and LiTFSI versus PS were studied with a combination of experimental and theoretical approaches. Figure 4A shows the appearances of the LiX/DME (X = DFTFSI or TFSI) electrolytes before and after the addition of Li2S6 solution as well as the reference solution of Li2S6/DME for comparison. Upon addition of Li2S6, the color of both LiDFTFSI- and LiTFSI-based electrolytes
Joule 3, 1689–1702, July 17, 2019 1695
immediately changes from colorless to brownish, being close to that of the reference Li2S6/DME solution, and the color remains unchanged after 60 h of stirring at room temperature (Figure 4A). The negligible difference in the appearance is confirmed by the ultraviolet-visible (UV-vis) measurement, where nearly superimposed traces of those three solutions in the visible light region (390–700 nm) are observed (Figure 4B). These results suggest the good chemical stability of both LiDFTFSI and LiTFSI toward PS species. Considering the possible chemical lability of H in the CF2H group, three types of reactions between PS and DFTFSI– are conceived and assessed with density functional theory (DFT) calculations. The corresponding Cartesian coordinates of the optimized structures are summarized in Table S3. As shown in Figure 4C, the abstraction of the H atom from the CF2H group with Li2S6 is the most unlikely reaction pathway (path 1) because of the lowest stabilization energy (DE) compared to the other two reactions (96 kJ mol1 [path 1] versus 178 kJ mol1 [path 2] versus 233 kJ mol1 [path 3]). In particular, the cleavage of the C–S bond from the CF3 side in DFTFSI– delivers a similar energy to that in DFTFSI–, being much lower than the PS reactive anion, bis(fluorosulfonyl)imide anion (FSI–, DEFSI = 654 kJ mol1).32 This suggests that the nucleophilic reaction of both LiDFTFSI and LiTFSI with Li2S6 are not energetically favorable, which is consistent with the results from visual and UV-vis tests in Figures 4A and 4B. Electrochemical Performance of Designer Anion in SSLSB Inspired by the aforementioned superior physicochemical and electrochemical properties of LiDFTFSI versus LiTFSI, the Li-S cells using both PEO-based SPEs were assembled and cycled at 70 C. Figures 5A and 5B show the discharge and charge profiles at different C-rates. The LiDFTFSI-based cell delivers a higher specific discharge and areal capacity than the LiTFSI-based one, e.g., 1,035 mAh g1 and 1.08 mAh cm2 (LiDFTFSI) versus 666 mAh g1 and 0.77 mAh cm2 (LiTFSI) for the third cycle at a charge and discharge rate of 0.05/0.05C. This could be ascribed to the higher Li-ion conductivity of LiDFTFSI-based electrolytes,33 which facilitates a faster Li-ion diffusion and migration in both the composite S electrode and bulk electrolyte, thereby bringing a higher degree of S utilization. As shown in Figure 5C, the LiDFTFSI-based cell exhibits high discharge capacity and cycling stability (i.e., CE close to 100%), whereas the LiTFSI-based one encounters PS shuttling after a few cycles as noted by the significantly declining CE upon cycling (e.g., CE = 74% for the 15th cycle at 0.2C and endless charging at the 34th cycle shown in Figure S7). The electrochemical impedance (EIS) spectra measured under fully charged state (Figure S8) show a sharper slope at the low-frequency region for the cycled LiTFSIbased cell than that for the LiDFTFSI-based one, further implying the uncompleted conversion of low-order PS (e.g., Li2S, Li2S2) species to elemental sulfur in the former cell.45 These results clearly suggest that the SEI layer on Li anode formed in LiDFTFSI-based electrolyte is not only dense and compact, which allows a highly reversible and dendrite-free Li plating and stripping, but also resistant enough toward the nucleophilic attack of the soluble PS intermediates. This corroborates well with the above mechanistic understanding of the electrochemical behavior of Li electrode in LiDFTFSI-based electrolyte (Figures 2 and 3). The benefit is such that the LiDFTFSI-based cell sustains more than 1,300 cycles (3,000 h continuous cycling) with CEs close to 100% after imposing the C-rate to a discharge and charge rate of 0.1/0.1C (Figure 5D), in spite of the reduced discharge and charge capacity due to increased cell polarization (Figure S9). Please note that no attempt was made to optimize the sulfur electrode (ratios, mixing) with the LiDFTFSI electrolyte. In effect, the gravimetric energy density and cyclability of the LiDFTFSI-based cell
1696 Joule 3, 1689–1702, July 17, 2019
Joule 3, 1689–1702, July 17, 2019 1697
Figure 5. Electrochemical Performance of Designer Anion in SSLSB (A and B) Discharge/charge profiles of the Li-S cells using LiX/PEO (X = DFTFSI [A] or TFSI [B]) electrolytes at 70 C. (C) Rate capability of the Li-S cells with LiX/PEO (X = DFTFSI or TFSI) electrolytes. (D) Long-term cycling stability of the Li-S cells using LiDFTFSI/PEO electrolytes at a discharge and charge rate of 0.1/0.1C (after C-rate test). (E) Comparison of the cycling performance for SSLSBs reported in literature and the results of this work. The numbers in squares correspond to the entries listed in Table S4, and color code corresponds to the type of electrolyte. The calculation details for the gravimetric energy density are available in Supplemental Information.
outperforms all reported results from literature, as summarized in Figure 5E. Those results suggest the structural design of Li salt offers a fertile ground for regulating the electrochemical performance of SSLSBs, as the anion chemistry governs (1) the ionic transport behavior in both bulk electrolyte and S cathode and (2) the composition and morphology of the SEI layer on Li anode. As seen in Figure S10, with a moderate S loading of ca. 1.0 mg cm2 , the LiDFTFSIbased cells show the best cell performance in terms of areal capacity (i.e., gravimetric energy density). Further increase of S loading leads to a significantly lower S utilization and a pronounced PS shuttling, thereby decreasing the energy efficiency and density at cell level (e.g., 1,035 mAh g3 [1.11 mgsulfur cm2] versus 502 mAh g1 [1.59 mgsulfur cm2]). Though the capacity retention and attainable energy density of the LiDFTFSI-based cells currently fall behind the stringent requirements of practical Li-S batteries, we believe that further optimization of electrolyte formation and precise control of cathode architecture could sufficiently improve the performance of the LiDFTFSI-based SSLSBs. For example, the replacement of PEO with a fully amorphous polymer matrix, such as Jeffamine-based polymers,38,46,47 could improve the ionic conductivity of SPEs in the room temperature region; the implementation of nano-structural design of S/C composite and cathode additives, which could increase the S loading and utilization and minimize the shuttling effect of PS intermediates, thereby improving the energy density of SSLSBs.48–50
DISCUSSION The designer anion, DFTFSI, with the substitution of CF3 moiety in the widely used TFSI with CF2H moiety, is proposed for coping with the sophisticated cell chemistry of Li-S batteries. In addition to the enhanced Li-ion conductivity due to the selective introduction of hydrogen atom in sulfonimide anion, LiDFTFSI shows several unprecedented features that are essential for high-performance SSLSBs: (1) the asymmetric structure of DFTFSI– and the presence of electrochemically labile CF2H moiety induce a facile decomposition of DFTFSI– on Li anode, leading to the formation of several critical SEI building species, such as ionic conductive LiH and mechanically stable LiF; (2) though a hydrogen atom is presented in the anionic structure, DFTFSI– is still chemically stable against the aggressive PS intermediates formed during the cycling of Li-S batteries, thus ensuring the good chemical compatibility of LiDFTFSI with the cell chemistry of Li-S batteries; (3) among all the reported results of SSLSBs (Table S4), the LiDFTFSI-based Li-S cells show ultrahigh gravimetric energy density and remarkable capacity retention upon prolonged cycling because of the enhanced Li-ion conductivity, high-quality SEI formed on Li anode, and good chemical stability of DFTFSI– versus PS intermediates. Further improvements of the electrolyte formation and cathode architectural design would sufficiently enhance the energy density of LiDFTFSI-based SSLSBs and thereby make such technology better than the state-of-the-art Li-S batteries employing liquid electrolytes. Together with the easy accessibility of LiDFTFSI in terms of synthesis, the LiDFTFSI-based electrolytes are promising candidates for improving
1698 Joule 3, 1689–1702, July 17, 2019
the electrochemical performances of SSLSBs, thus promoting their practical deployment in the future.
EXPERIMENTAL PROCEDURES Preparation of Polymer Electrolytes LiDFTFSI was synthesized according to our recent work.33,34 Polymer membranes with an average thickness of 50 mm were prepared by conventional solvent casting method, followed by hot-pressing (High Temperature Film Maker Controller, Specac).33,34 The polymer was dissolved into acetonitrile, and then the salts, either LiDFTFSI or LiTFSI (battery grade, Solvionic, France), were added. In all cases, the salt concentration was 20 (the molar ratio of CH2CH2O [EO]/Li+). Electrochemical Stability of Electrolyte/Li Interphase Li symmetrical coin cells (Li jj Li ) were assembled in an argon-filled glove box to investigate the electrochemical stability between electrolyte and Li interphase. The galvanostatic cycling of the Li symmetric cells was evaluated using a VMP 3 potentiostat (Bio-Logic Science Instruments). The Li symmetric cells were cycled galvanostatically at a current density of 0.1 mA cm2, wherein the duration of each half-cycle was 2 h. Surface Morphology and Composition of the Cycled Li Electrode Surface morphologies of the Li deposits were examined by a field emission Quanta 200 FEG (FEI) microscope, operated at 20 kV. The Li deposits were obtained by the galvanostatic deposition of Li on Cu substrates using Li j LiX/DME (X = DFTFSI or TFSI) j Cu cells at a current density of 0.1 mA cm2 for 20 h. The compositions of the surface layer were measured by a Phoibos 150 XPS with a non-monochromatic Mg Ka source (hn = 1,253.6 eV). The spectra were recorded with high-resolution scans at low power (100 W, 20 eV pass energy, and 0.1 eV energy step). The calibration of the binding energy was performed taking into account as reference the Auger parameter of LiF at 1,340 eV.51 The samples were gently rinsed with DME and dried thoroughly under vacuum before transferred to the SEM or XPS chamber by a homedesigned airtight setup. Electrochemical Reduction Simulation Naphthalene was dissolved in ultra-pure and dry THF for about 30 min. Following, ground Li was added to the naphthalene and THF solution in a 1:1 mole ratio (naphthalene: Li). The reaction mixture was stirred for about 4 h at room temperature using a special glass-coated magnetic stirring bar as polytetrafluoroethylene (PTFE) is reduced at these potentials. Afterward, a dark green color, characteristic of the radical anion was observed (Figure S2). Following the complete formation of the radical anion, the reduction of both LiDFTFSI and LiTFSI was tested by adding the corresponding amounts of the salt. Finally, the residual color was monitored. Reaction of Polysulfide with Anions The solution of 0.1 M Li2S6/DME had been prepared by mixing Li2S and S with a molar ratio of 1:5 in DME, followed by stirring for 1 week at room temperature.42 The possible reaction between PS species and the two investigated anions were carried out in DME solution at room temperature. To a stirred solution of 1 M LiX/DME (X = DFTFSI or TFSI), a predetermined amount of 0.1 M Li2S6/DME solution with a molar ratio of LiX/Li2S6 of 200 was added. The UV-vis measurements were carried out, without dilution, on a Cary 5000 UV-vis spectrophotometer (Varian).
Joule 3, 1689–1702, July 17, 2019 1699
Our DFT calculations were based on the Becke’s three parameters (B3) exchange functional along with the Lee-Yang-Parr (LYP) non-local correlation functional (B3LYP),52,53 as implemented in the numeric atom-centered basis set all-electron code FHI-aims.54,55 The starting geometries of the investigated sulfonamide anions, considering different conformers, were generated by varying either S‒N‒S‒R1 or S‒N‒S‒R2 dihedral angles. The initial geometries were constructed using the open-source molecular editor and visualizer Avogadro.56 We used the universal force field and genetic algorithm search tool as implemented in Avogadro to prescreen dozens of low-energy geometries. Then, their atomic structures using a trust-radius-method-enhanced version of the Broyden-Fletcher-Goldfarb-Shanno optimization algorithm57 were fully relaxed. We used ‘‘tight’’ settings and ‘‘tier2’’ standard basis sets for H, Li, C, N, O, F, and S atoms. A maximum residual force component per atom of 0.01 eV A˚1 was the threshold for the convergence criteria in all structural relaxations, together with 104 electrons for the electron density and 106 eV for the total energy of the system. S Cathode Preparation and Cycling of Li-S Polymer Cell Composite sulfur cathode was prepared with 40 wt % elemental sulfur (99.5 wt %, Sigma-Aldrich), 15 wt % conductive carbons (Ketjen Black, KJ600, Akzo-Nobel), and 45 wt % LiX/PEO (X = DFTFSI or TFSI) as electrolyte. The S loading was from 1.0 to 1.1 mg cm2. The procedure has been detailed in our previous work.30,31 Li-S polymer cells were assembled in an argon-filled glovebox using the prepared electrode as cathode, polymer electrolyte membrane as both electrolyte and separator, and Li disk (China Energy Lithium) as anode. The cells were then cycled galvanostatically at a constant current mode between 1.6 and 2.8 V at 70 C using a BT lab cycler (Bio-Logic Science Instruments). Meanwhile, the EIS spectra of the cells in the fully charged state after various cycles were collected over the frequencies from 104 to 102 Hz using the above BT lab cycler.
SUPPLEMENTAL INFORMATION Supplemental Information can be found online at https://doi.org/10.1016/j.joule. 2019.05.003.
ACKNOWLEDGMENTS This work was supported by the Ministerio de Economı´a y Competitividad (MINECO) of the Spanish Government through Proyectos I + D Retos program (ENE201564907C2-1-R and ENE2016-81020-R grants). We also acknowledge funding by the Basque Government through the GVELKARTEK-2016 program. X.J. thanks the Basque Government for PhD funding, C.L. thanks the Spanish Government for the Juan de la Cierva scholarship (ref: FJCI-2015-23898), and H.Z. thank the Basque Government for the Berrikertu program (1-AFW-2017-2). We are also grateful for computer resources to the i2BASQUE academic network and SGI/IZOSGIker UPV/ EHU (Arina cluster) and the XPS measurements and discussions with Dr. Miguel A´ngel Mun˜oz-Ma´rquez. We thank Solvay, especially Claude Mercier, for the generous supply of difluoromethanesulfonyl chloride.
AUTHOR CONTRIBUTIONS H.Z. and M.A. conceived the research and designed the experiments. U.O., X.J., G.G.E., M.M.-I., and C.L. carried out the experiments and measurements. H.Z. and J.C. performed the computational work. All authors discussed the analysis of results. H.Z. and M.A. wrote the manuscript with input from all authors. H.Z. and M.A. supervised the work.
1700 Joule 3, 1689–1702, July 17, 2019
DECLARATION OF INTERESTS The authors declare no competing interests. Received: March 6, 2019 Revised: April 19, 2019 Accepted: May 3, 2019 Published: May 24, 2019
REFERENCES 1. Smalley, R.E. (2005). Future Global Energy Prosperity: The Terawatt Challenge. MRS Bulletin 30, 412–417.
16. Cao, R., Xu, W., Lv, D., Xiao, J., and Zhang, J.-G. (2015). Anodes for rechargeable lithium-sulfur batteries. Adv. Energy Mater 5, 1402273.
2. Dunn, B., Kamath, H., and Tarascon, J.M. (2011). Electrical energy storage for the grid: a battery of choices. Science 334, 928–935.
17. Lin, D., Liu, Y., and Cui, Y. (2017). Reviving the lithium metal anode for high-energy batteries. Nat. Nanotechnol 12, 194–206.
3. Tarascon, J.M., and Armand, M. (2001). Issues and challenges facing rechargeable lithium batteries. Nature 414, 359–367.
18. Cheng, X.B., Zhang, R., Zhao, C.Z., and Zhang, Q. (2017). Toward safe lithium metal anode in rechargeable batteries: a review. Chem. Rev. 117, 10403–10473.
4. Armand, M., and Tarascon, J.M. (2008). Building better batteries. Nature 451, 652–657. 5. Nitta, N., Wu, F., Lee, J.T., and Yushin, G. (2015). Li-ion battery materials: present and future. Mater. Today 18, 252–264. 6. Goodenough, J.B. (2018). How we made the Liion rechargeable battery. Nat. Electron 1, 204. 7. Cano, Z.P., Banham, D., Ye, S., Hintennach, A., Lu, J., Fowler, M., and Chen, Z. (2018). Batteries and fuel cells for emerging electric vehicle markets. Nat. Energy 3, 279–289. 8. Bruce, P.G., Freunberger, S.A., Hardwick, L.J., and Tarascon, J.M. (2011). Li-O2 and Li-S batteries with high energy storage. Nat. Mater 11, 19–29. 9. Choi, J.W., and Aurbach, D. (2016). Promise and reality of post-lithium-ion batteries with high energy densities. Nat. Rev. Mater 1, 16013. 10. Zhang, H., Eshetu, G.G., Judez, X., Li, C., Rodriguez-Martı´nez, L.M., and Armand, M. (2018). Electrolyte additives for lithium metal anodes and rechargeable lithium metal batteries: progresses and perspectives. Angew. Chem. Int. Ed. 57, 15002–15027. 11. Shen, Y., Zhang, Y., Han, S., Wang, J., Peng, Z., and Chen, L. (2018). Unlocking the energy capabilities of lithium metal electrode with solid-state electrolytes. Joule 2, 1674–1689. 12. Judez, X., Eshetu, G.G., Li, C., RodriguezMartinez, L.M., Zhang, H., and Armand, M. (2018). Opportunities for rechargeable solidstate batteries based on Li-intercalation cathodes. Joule 2, 2208–2224. 13. Yin, Y.X., Xin, S., Guo, Y.G., and Wan, L.J. (2013). Lithium-sulfur batteries: Electrochemistry, materials, and prospects. Angew. Chem. Int. Ed. Engl. 52, 13186–13200. 14. Nazar, L.F., Cuisinier, M., and Pang, Q. (2014). Lithium-sulfur batteries. MRS Bull. 39, 436–442. 15. Xu, W., Wang, J., Ding, F., Chen, X., Nasybulin, E., Zhang, Y., and Zhang, J.-G. (2014). Lithium metal anodes for rechargeable batteries. Energy Environ. Sci. 7, 513–537.
19. Fenton, D.E., Parker, J.M., and Wright, P.V. (1973). Complexes of alkali metal ions with poly(ethylene oxide). Polymer 14, 589. 20. Armand, M., Chabagno, J.M., and Duclot, M.J. (1978). Second International Meeting on Solid Electrolytes Extended Abstracts. St. Andrews Ecosse. September 20–22. 21. Armand, M.B. (1986). Polymer electrolytes. Annu. Rev. Mater. Sci. 16, 245–261. 22. Di Noto, V., Lavina, S., Giffin, G.A., Negro, E., and Scrosati, B. (2011). Polymer electrolytes: present, past and future. Electrochim. Acta 57, 4–13. 23. Hallinan, D.T., and Balsara, N.P. (2013). Polymer electrolytes. Annu. Rev. Mater. Res. 43, 503–525. 24. Zhang, H., Li, C., Piszcz, M., Coya, E., Rojo, T., Rodriguez-Martinez, L.M., Armand, M., and Zhou, Z. (2017). Single lithium-ion conducting solid polymer electrolytes: advances and perspectives. Chem. Soc. Rev. 46, 797–815. 25. Li, C., Zhang, H., Otaegui, L., Singh, G., Armand, M., and Rodriguez-Martinez, L.M. (2016). Estimation of energy density of Li-S batteries with liquid and solid electrolytes. J. Power Sources 326, 1–5. 26. Judez, X., Zhang, H., Li, C., Eshetu, G.G., Gonza´lez-Marcos, J.A., Armand, M., and Rodriguez-Martinez, L.M. (2018). Review— Solid electrolytes for safe and high energy density lithium-sulfur batteries: promises and challenges. J. Electrochem. Soc 165, A6008– A6016. 27. Xu, K. (2014). Electrolytes and interphases in Liion batteries and beyond. Chem. Rev. 114, 11503–11618. 28. Younesi, R., Veith, G.M., Johansson, P., Edstro¨m, K., and Vegge, T. (2015). Lithium salts for advanced lithium batteries: Li–metal, Li–O2, and Li–S. Energy Environ. Sci. 8, 1905–1922. 29. Zhang, H., Liu, C., Zheng, L., Xu, F., Feng, W., Li, H., Huang, X., Armand, M., Nie, J., and Zhou, Z. (2014). Lithium bis(fluorosulfonyl)imide/
poly(ethylene oxide) polymer electrolyte. Electrochim. Acta 133, 529–538. 30. Judez, X., Zhang, H., Li, C., Gonza´lez-Marcos, J.A., Zhou, Z., Armand, M., and RodriguezMartinez, L.M. (2017). Lithium bis(fluorosulfonyl) imide/poly(ethylene oxide) polymer electrolyte for all solid-state Li–S cell. J. Phys. Chem. Lett. 8, 1956–1960. 31. Eshetu, G.G., Judez, X., Li, C., Bondarchuk, O., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2017). Lithium azide as an electrolyte additive for all-solid-state lithium– sulfur batteries. Angew. Chem. Int. Ed. Engl. 56, 15368–15372. 32. Eshetu, G.G., Judez, X., Li, C., Martinez-Iban˜ez, M., Gracia, I., Bondarchuk, O., Carrasco, J., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2018). Ultrahigh performance all solid-state lithium sulfur batteries: salt anion’s chemistry-induced anomalous synergistic effect. J. Am. Chem. Soc. 140, 9921–9933. 33. Zhang, H., Oteo, U., Zhu, H., Judez, X., Martinez-Iban˜ez, M., Aldalur, I., Sanchez-Diez, E., Li, C., Carrasco, J., Forsyth, M., et al. (2019). Enhanced lithium-ion conductivity of polymer electrolytes by selective introduction of hydrogen into the anion. Angew. Chem. Int. Ed. Engl. https://doi.org/10.1002/ange. 201813700. 34. Oteo, U., Martinez-Iban˜ez, M., Aldalur, I., Sanchez-Diez, E., Carrasco, J., Armand, M., and Zhang, H. (2019). Improvement of the cationic transport in polymer electrolytes with (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide salts. ChemElectroChem 6, 1019–1022. 35. Sun, X.-G., and Kerr, J.B. (2006). Synthesis and characterization of network single ion conductors based on comb-branched polyepoxide ethers and lithium bis(allylmalonato)borate. Macromolecules 39, 362–372. 36. Ikeya, M. (1977). Electrical properties of lithium hydride. J. Phys. Soc. Jpn 42, 168–174. 37. Lin, Z., Liu, Z., Dudney, N.J., and Liang, C. (2013). Lithium superionic sulfide cathode for all-solid lithium–sulfur batteries. ACS Nano 7, 2829–2833. 38. Aldalur, I., Martinez-Iban˜ez, M., Piszcz, M., Zhang, H., and Armand, M. (2018). Selfstanding highly conductive solid electrolytes based on block copolymers for rechargeable all-solid-state lithium-metal batteries. Batteries & Supercaps 1, 149–159. 39. Judez, X., Zhang, H., Li, C., Eshetu, G.G., Zhang, Y., Gonza´lez-Marcos, J.A., Armand, M.,
Joule 3, 1689–1702, July 17, 2019 1701
and Rodriguez-Martinez, L.M. (2017). Polymer-rich composite electrolytes for all-solid-state Li–S cells. J. Phys. Chem. Lett. 8, 3473–3477. 40. Qian, J., Henderson, W.A., Xu, W., Bhattacharya, P., Engelhard, M., Borodin, O., and Zhang, J.G. (2015). High rate and stable cycling of lithium metal anode. Nat. Commun 6, 6362. 41. Aurbach, D., Daroux, M.L., Faguy, P.W., and Yeager, E. (1988). Identification of surface films formed on lithium in dimethoxyethane and tetrahydrofuran solutions. J. Electrochem. Soc. 135, 1863–1871. 42. Suo, L., Hu, Y.S., Li, H., Armand, M., and Chen, L. (2013). A new class of solvent-in-salt electrolyte for high-energy rechargeable metallic lithium batteries. Nat. Commun 4, 1481. 43. Eshetu, G.G., Diemant, T., Grugeon, S., Behm, R.J., Laruelle, S., Armand, M., and Passerini, S. (2016). In-depth interfacial chemistry and reactivity focused investigation of lithium– imide- and lithium–imidazole-based electrolytes. ACS Appl. Mater. Interfaces 8, 16087–16100. 44. Aurbach, D., and Weissman, I. (1999). On the possibility of LiH formation on Li surfaces in wet electrolyte solutions. Electrochem. Commun 1, 324–331. 45. Deng, Z., Zhang, Z., Lai, Y., Liu, J., Li, J., and Liu, Y. (2013). Electrochemical impedance
1702 Joule 3, 1689–1702, July 17, 2019
spectroscopy study of a lithium/sulfur battery: modeling and analysis of capacity fading. J. Electrochem. Soc. 160, A553–A558. 46. Aldalur, I., Martinez-Iban˜ez, M., Piszcz, M., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2018). Lowering the operational temperature of all-solid-state lithium polymer cell with highly conductive and interfacially robust solid polymer electrolytes. J. Power Sources 383, 144–149. 47. Aldalur, I., Martinez-Iban˜ez, M., Krzto nMaziopa, A., Piszcz, M., Armand, M., and Zhang, H. (2019). Flowable polymer electrolytes for lithium metal batteries. J. Power Sources 423, 218–226. 48. Ma, J., Fang, Z., Yan, Y., Yang, Z., Gu, L., Hu, Y.-S., Li, H., Wang, Z., and Huang, X. (2015). Novel large-scale synthesis of a C/S nanocomposite with mixed conducting networks through a spray drying approach for Li–S batteries. Adv. Energy Mater 5, 1500046. 49. Yang, Y., Zheng, G., and Cui, Y. (2013). Nanostructured sulfur cathodes. Chem. Soc. Rev. 42, 3018–3032. 50. Pang, Q., Shyamsunder, A., Narayanan, B., Kwok, C.Y., Curtiss, L.A., and Nazar, L.F. (2018). Tuning the electrolyte network structure to invoke quasi-solid state sulfur conversion and suppress lithium dendrite formation in Li–S batteries. Nat. Energy 3, 783–791.
51. Wagner, C.D., Riggs, W.M., Davis, L.E., Moulder, J.F., and Muilenberg, G.E. (1979). A Reference Book of Standard Data for Use in X-Ray Photoelectron Spectroscopy (PerkinElemer). 52. Lee, C., Yang, W., and Parr, R.G. (1988). Development of the Colle-Salvetti correlationenergy formula into a functional of the electron density. Phys. Rev. B Condens. Matter 37, 785–789. 53. Becke, A.D. (1993). Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98, 5648–5652. 54. Blum, V., Gehrke, R., Hanke, F., Havu, P., Havu, V., Ren, X., Reuter, K., and Scheffler, M. (2009). Ab initio molecular simulations with numeric atom-centered orbitals. Comput. Phys. Commun 180, 2175–2196. 55. Havu, V., Blum, V., Havu, P., and Scheffler, M. (2009). Efficient integration for all-electron electronic structure calculation using numeric basis functions. J. Comput. Phys. 228, 8367– 8379. 56. Hanwell, M.D., Curtis, D.E., Lonie, D.C., Vandermeersch, T., Zurek, E., and Hutchison, G.R. (2012). Avogadro: an advanced semantic chemical editor, visualization, and analysis platform. J. ChemInform 4, 17. 57. Nocedal, J., and Wright, S.J. (2006). Sequential Quadratic Programming (Springer).
Designer Anion Enabling Solid-State LithiumSulfur Batteries Heng Zhang, Uxue Oteo, Xabier Judez, ..., Javier Carrasco, Chunmei Li, Michel Armand [email protected] (H.Z.) [email protected] (M.A.)
HIGHLIGHTS Scalable designer anion is proposed for solid-state Li-S batteries (SSLSBs) New designer anion shows superior stability against Li metal electrode Long-term cycling of SSLSBs is achieved with designer anion Structural modification of salt anions tunes the performance of Li metal batteries
In this work, we developed a new designer anion favoring Li-ion transport and improving interfacial compatibility with Li metal electrode, thus boosting the performance of solid-state Li-S batteries (SSLSBs). The designer anion could be made with the industrially available intermediates and is believed to expedite the development of SSLSBs as well as other rechargeable Li batteries.
Zhang et al., Joule 3, 1689–1702 July 17, 2019 ª 2019 Elsevier Inc. https://doi.org/10.1016/j.joule.2019.05.003
Article
Designer Anion Enabling Solid-State Lithium-Sulfur Batteries Heng Zhang,1,* Uxue Oteo,1 Xabier Judez,1 Gebrekidan Gebresilassie Eshetu,2,3 Maria Martinez-Iban˜ez,1 Javier Carrasco,1 Chunmei Li,1 and Michel Armand1,4,*
SUMMARY
Context & Scale
With an extremely high theoretical energy density, solid-state lithium-sulfur (Li-S) batteries (SSLSBs) are emerging as one of the most feasible chemistries; however, their energy efficiency and long-term cyclability are severely hampered by the lithium metal (Li ) dendrite formation during repeated discharge/charge cycles and the shuttling of aggressive polysulfide intermediates between two electrodes. Herein, we report (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide anion [N(SO2CF2H)(SO2CF3)], hereafter DFTFSI, as a designer anion for high-performance polymer-based SSLSBs. In contrast to the widely used bis(trifluoromethanesulfonyl)imide anion [N(SO2CF3)2] (TFSI), DFTFSI-based SSLSBs provide superior interfacial stability against Li , extremely high discharge and areal capacities, very high Coulombic efficiency, and long-term cyclability, surpassing the reported literature values, in terms of gravimetric energy density. This work opens a new door for accelerating the practical deployment of SSLSBs in the future.
Solid-polymer-electrolyte (SPE)based solid-state Li-S batteries (SSLSBs) are endowed with profuse beneficial futures such as high theoretical energy density, low cost, higher safety, and environmental benignancy. However, their practical deployment is hindered by several intrinsic problems, e.g., Li metal (Li ) dendrite or mossy growth, electronically insulating nature of S8 and Li2S, polysulfides shuttling, etc. This work presents a breakthrough in improving the performance of SSLSBs via a designer anion, (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide anion (DFTFSI). DFTFSI-based electrolytes show dendrite-free Li plating and stripping, and enable the long-term cycling of LiS cells with high capacities and excellent Coulombic efficiencies. Thus, this study opens a new avenue toward the design of new and tailored SPEs for applications in a high-performance and safer Li-S battery as well as other rechargeable Li batteries.
INTRODUCTION Sustainable development of today’s society puts forward an increasing demand for efficient and clean usage of energy sources.1,2 Lithium (Li)-ion batteries (LIBs), an important part for building the electrified network, have been widely deployed as power sources for smart electronic and (hybrid) electric vehicles (xEVs), as well as buffers for intermittent renewable energy (e.g., wind, solar).3–7 With a ceiling gravimetric energy of ca. 400 Wh kg1, LIBs still fall behind the requirement of future xEVs and other emerging applications.8–12 Among all the post Li-ion chemistries, Li-sulfur (Li-S) batteries appear to be the most appealing and viable energy storage technologies in view of their extraordinary energy density (theoretical value of ca. 2,600 Wh kg1), abundant sulfur resources, cost effectiveness, and environmental benignancy.8,13,14 However, the practical implementation of Li-S batteries faces formidable challenges, such as the low electronic conductivity of elemental sulfur (S8) and Li sulfide (Li2S), shuttling effect of the reaction intermediates (i.e., polysulfide [PS]), and the notorious Li metal (Li ) dendrite formation bearing along low Coulombic efficiencies and safety concerns in the presence of volatile liquid ethers solution.8,13–18 Solid polymer electrolytes (SPEs), first found by Wright and coworkers19 in 1973 and then suggested by Armand20 as electrolytes for rechargeable batteries in 1978, have been recognized as promising candidates for accessing safe and high-performance rechargeable Li -based batteries.21–24 This is motivated by their intrinsically enhanced safety compared to conventional liquid electrolytes, good processability, and flexibility compared to inorganic solid electrolytes.21–24 Besides, the
Joule 3, 1689–1702, July 17, 2019 ª 2019 Elsevier Inc. 1689
replacement of liquid electrolytes with SPEs could remarkably enhance the energy density of Li-S batteries as shown in recent calculations.25,26 As one of the most critical parameters, the nature of salt anions plays a pivotal role in dictating the electrochemical performance of solid-state Li-S batteries (SSLSBs).26–28 To date, bis(trifluoromethanesulfonyl)imide anion {[N(SO2CF3)2], TFSI} has captured the most interest among all the anions evaluated for SSLSBs because of its good thermal and chemical stability, structural flexibility, and plasticizing effect when coupled with the widely used poly(ethylene oxide) (PEO) matrix.21–24 However, the solid-electrolyte interphase (SEI) layers formed between the LiTFSI/PEO electrolyte and Li anode are not stable enough toward the aggressive PS intermediates, thereby resulting in an inferior cycling performance of SSLSBs.29–32 The (fluorosulfonyl)(trifluoromethanesulfonyl)imide anions {[N(SO2F)(SO2CF3)], FTFSI} revealed in our recent work clearly shows the importance of designer anion on the electrochemical performance of SSLSBs.32 The corresponding Li-S cells delivered high specific and areal capacity with good Coulombic efficiency (CE) because of the synergistic effects resulting from its molecular structure (i.e., possessing both SO2CF3 and SO2F functionalities). However, LiFTFSI is susceptible to water, aprotic solvents, and heat, limiting the scalable preparation of its electrolytes, and the stability of the Li electrode still needs to be improved for long-term cycling of SSLSBs. In this work, we demonstrate that a designer anion, (difluoromethanesulfonyl)(trifluoromethanesulfonyl)imide {[N(SO2CF2H)(SO2CF3)], DFTFSI}, hoists the performance of polymer-based SSLSBs. The chemical structure of DFTFSI is designed on the basis of the following considerations (schematically illustrated in Figure 1). Firstly, the CF2H moiety, being weakly acidic, is beneficial for restricting anionic mobility via the hydrogen-bonding interaction between DFTFSI– and PEO, resulting in an enhanced Li-ion conductivity (sLi+ = 2 3 104 S cm1 [LiDFTFSI/PEO] versus sLi+ = 1.5 3 104 S cm1 [LiTFSI/PEO] at 70 C33,34). Of note is that higher anionic mobility is responsible for the steeper concentration gradient accumulated during the discharge and charge processes, which causes an increased polarization, voltage loss, and cell failure of SSLSBs.15,24,35 Secondly, the asymmetric structure of DFTFSI– and the presence of electrochemically labile CF2H moiety might induce a facile decomposition of DFTFSI– on a Li anode, leading to the formation of several important SEI building species, such as ionic conductive Li hydride (LiH [s = 10–10 S cm–1]36 versus Li2S [s = 10–13 S cm–1]37 at 25 C) and mechanically stable Li fluoride. Lastly, the preparation of DFTFSI– could be made in a sequence of high-yield reactions with the industrially available intermediates (e.g., difluoromethanesulfonyl chloride from Solvay), thus ensuring the potential large-scale application of LiDFTFSI in SSLSBs.
RESULTS The Chemistry of Designer Anion with Li Electrode Firstly, the effect of salt concentration on the phase-transition behavior and ionic conductivity of LiDFTFSI/PEO is investigated prior to the understanding of anion chemistry on Li electrode. As seen in Figure S1, at the mole ratio of ethylene oxide (EO)/Li = 20, LiDFTFSI/PEO shows a low glass transition temperature (Tg) at 38 C. With the addition of Li salt, the value of Tg increases gradually up to 30 C for the sample of EO/Li = 8. As a consequence, the ionic conductivity of LiDFTFSI/PEO shows the highest at EO/Li = 20 and drops at a higher salt content (see Figure S2). Therefore, LiDFTFSI- and LiTFSI-based electrolytes at EO/Li = 20 were selected for the following electrochemical studies. In addition, as displayed in Figure S3, the
1690 Joule 3, 1689–1702, July 17, 2019
1Electrical
Energy Storage Department, CIC Energigune, Parque Tecnolo´gico de A´lava, Albert Einstein 48, 01510 Min˜ano, A´lava, Spain 2Institute
for Power Electronics and Electrical Drives (ISEA), RWTH Aachen University, Ja¨gerstraße 17/19, 52066 Aachen, Germany
3Department
of Chemistry, College of Natural and Computational Sciences, Mekelle University, P.O. Box 231, Mekelle, Ethiopia
4Lead
Contact
*Correspondence: [email protected] (H.Z.), [email protected] (M.A.) https://doi.org/10.1016/j.joule.2019.05.003
Figure 1. The Role of LiDFTFSI and LiTFSI in SSLSBs
anodic stability of LiDFTFSI/PEO and LiTFSI/PEO is higher than 4 V versus Li/Li+, which is well acceptable for the operation of Li-S batteries. Highly reversible and dendrite-free plating and stripping of Li electrode is crucial for the long-term cyclability, rate capability, CE, and safety of SSLSBs. The electrochemical stabilities of Li electrode in both LiDFTFSI- and LiTFSI-based electrolytes are comparatively displayed in Figure 2. As shown in Figure 2A, the galvanostatic cycling of Li symmetric cells using the LiDFTFSI/PEO electrolyte sustains more than 600 h without any erratic voltage oscillation, whereas the LiTFSI-based one encounters an internal short circuit after 168 h of cycling.38 The excellent stability of Li electrode in the LiDFTFSI-based electrolyte is further manifested by the superior cycling stability of Li symmetric cells using the corresponding liquid electrolyte, where the cycle life of Li j LiDFTFSI/1,2-dimethoxyethane (DME) j Li cell is 25 times higher than that of the LiTFSI-based one (Figure 2B, > 1,000 h [LiDFTFSI/DME] versus ca. 40 h [LiTFSI/DME]32). The markedly enhanced cyclability of the LiDFTFSI-based cell strongly evidences the formation of stable SEI layers on Li electrode. In addition, the slightly higher overvoltage in the LiDFTFSI-based cells (both liquid and polymer electrolytes) than in LiTFSI-based ones implies that the SEI layers generated in the former electrolytes tend to be more resistive and compact (see Figures 2D and 2E for X-ray photoelectron spectroscopy [XPS] spectra). To elucidate the role of CF2H moiety on the electrochemical behavior of Li electrode, post-mortem analyses employing scanning electron microscopy (SEM) and XPS were performed in the liquid cell configuration. Note that Li electrode in contact with PEO-based electrolytes cannot be separated without damaging its surface because of the strong adhesive properties of PEO; hence, DME having an analogous chemical structure to PEO was used for assessing the Li j electrolyte interphase.30–32,39 As seen in Figure 2C, the Li deposits on copper substrate (2 mAh cm2 deposition) recovered from the LiDFTFSI-based electrolyte show
Joule 3, 1689–1702, July 17, 2019 1691
Figure 2. Electrochemical Compatibility of Designer Anion with Li Electrode (A and B) Galvanostatic cycling of Li electrode in the PEO-based electrolytes at 70 C (A) and DMEbased electrolytes at 25 C (B). Current density, 0.1 mA cm2 ; half-cycle time, 2 h. (C) Physical appearances and SEM images of Li deposited onto Cu substrates at 0.1 mA cm 2 (plating time 20 h) in LiX/DME (X = DFTFSI or TFSI) electrolytes at 25 C. (D) XPS spectra of the corresponding Li deposits at 25 C. In the figure, R 1 CO and R 2 CO refer to CH3 O and H2 C=HCOCH 3 , respectively; and the black and violet lines correspond to the respective raw and fitted data. (E) Atomic concentration of Li, C, O, F, S, and N on the Li deposits with different etching time using the two liquid electrolytes. For comparison, the raw data of Li plating and stripping and XPS spectra of LiTFSI sample are retreated and presented from Aldalur et al. 38 and Eshetu et al.32 with permission. Copyright 2018 Wiley-VCH Verlag GmbH & Co. KGaA; Copyright 2018 American Chemical Society.
better coverage and homogeneity than the ones from the LiTFSI-based electrolyte, presumably because of the enhanced kinetics for the Li nucleation and growth steps. Besides, smooth, uniform, and pancake-like morphologies of Li deposits are observed for the LiDFTFSI-based electrolyte. In comparison, a large amount of typical needle-like Li dendrites are formed in the LiTFSI-based one.31,32,40 This distinct morphological observation between LiDFTFSI and LiTFSI-based electrolytes clearly suggests that the replacement of CF3 group in TFSI– with a CF2H group
1692 Joule 3, 1689–1702, July 17, 2019
endows the formation of compact and dense Li deposits that are critical for the long-term cycling of SSLSBs. Figure 2D presents the most relevant spectra, C1s and F1s, of the outermost surface layer covered on Li deposits. The assignments of each peak in the XPS spectra as extracted in the previous reports are summarized in Table S1, and the corresponding survey spectra are given in Figure S4. The C1s spectra of both electrolyte systems show dominative peaks in the range of 289 to 284 eV, which are attributed to the presence of alkoxide (e.g., CH3OLi, R1CO) and methoxy-vinyl-ether (e.g., H2C=HCOCH3, R2CO) resulting from the electrochemical reduction of DME on the Li surface. The reduction mechanism of DME has been extensively discussed in our recent work32 as well as previous work from Aurbach et al.41 and Hu et al.42 (Scheme S1), where the scission of the CO bond in DME could form SEI building species such as RCOLi and RCOCR. In addition, the minor signals at slightly higher binding energies (i.e., 294–290 eV) belong to the salt-derived species, such as CF3- and CF2-containing components. Those species are also observed in the corresponding F1s spectra as shown in Figure 2D. Notably, the F1s signal at 685 eV associated with LiF is more prominent for the LiDFTFSI-based electrolyte than that for the LiTFSI-based one (see Table S2 for the normalized percentage of each F-containing species). This suggests that the electrochemical reduction of LiDFTFSI is easier and more exhaustive than that of LiTFSI because of its asymmetric structure and the presence of an electrochemically labile CF2H group. The less negative reduction potential (Ere) for LiDFTFSI is further demonstrated by the chemical simulation experiments with naphthalene radical mono-anions as reducing agents. As shown in Figure S5, LiDFTFSI gets reduced easily upon the addition of naphthalene radical anions (ca. 1 versus Li/Li+43)/tetrahydrofuran (THF) solution, whereas LiTFSI remains stable under the same condition. Previous work revealed that the reduction of LiTFSI is difficult even when treated with biphenyl di-radical anions, which has a potential of ca. 0.15 V versus Li/Li+.31,43 To ascertain the nature and properties of the SEI layer formed in the LiDFTFSI- and LiTFSI-based electrolytes, depth-dependent compositions of several typical elements (Li, C, O, F, S, and N) are shown in Figure 2E. For both electrolytes, the Li concentration increases with increase of the sputtering time with Ar+ ions because of the removal of outermost reduction products (e.g., Li salts and solvents) on the SEI layer. Yet, the Li concentration of the LiDFTFSI-based electrolyte reaches a constant value in a shorter sputtering time (120s [LiDFTFSI] versus 540s [LiTFSI]), suggesting a compact and dense SEI layer formed in LiDFTFSI-based electrolyte. Meanwhile, the F1s spectra gathered from various sputtering times (Figure S6) show that LiF, the ultimate reduction product, becomes a dominant component at the inner part of the formed SEI layer after only 30s sputtering for the LiDFTFSI-based electrolyte. This again proves that the electrochemical reduction of LiDFTFSI promotes the formation of a thin, LiF-rich SEI layer. As shown in Scheme S1, the reduction of LiDFTFSI possibly leads to the formation of ionic conductive SEI building species, LiH, though it could be observed from neither XPS spectra because of the overlapping between LiH and LiOH (unavoidably present as native SEI species on Li electrode) nor Fourier transform infrared spectroscopy (FTIR) spectra as carried out by Aurbach et al.44 because of the concurrent FTIR absorptions of LiH, and SO2 moieties in sulfonimide salts and CO in ethereal solvents. Thus, the combination of mechanically stable LiF and ionic conductive LiH as SEI building materials in the LiDFTFSI-based electrolyte largely inhibits the Li dendrite growth and thereby improves the electrochemical stability of Li electrode.
Joule 3, 1689–1702, July 17, 2019 1693
Figure 3. Mechanistic Understanding of Designer Anion with Li Electrode (A and B) Proposed mechanism for the reduction of the LiDFTFSI (A) and LiTFSI (B) salts. (C) Schematic illustration on the SEI layer formed on Li electrode in both salts.
On the basis of the above-mentioned morphological (Figure 2C) and compositional information (Figures 2D and 2E) of Li electrode surface, together with the chemical reduction simulation tests of the neat Li salts, the SEI layers generated in the LiDFTFSI- and LiTFSI-based electrolytes are schematically depicted in Figure 3. An initial two-electron reduction of DFTFSI– and TFSI– gives alkyl sulfonate or alkyl sulfonylamide intermediates (Figures 3A and 3B), e.g., CF2QSO2Li and CF2QSO2NLi2 (Q = H or F) for DFTFSI–. Sequential reduction of the intermediates affords electrochemically stable inorganic species, such as LiF, LiH, Li2S, and Li3N. Alternatively, these highly reductive species are also accessible with the rigorous degradation of the anions on Li surface, which processes the lowest possible electrochemical potential (–3.040 V versus standard hydrogen electrode [SHE]15). As seen in Figure 3C, the SEI layers covered on Li electrode consist of three parts, including (1) the outermost organic-rich layer emanating from solvent reduction and/or residual salts trapped in the passivation layer, (2) inorganic-rich layer derived from the decomposition of Li salts, and (3) innermost native SEI layer inherited from the Li production and processing. The distinct difference between two electrolytes on the SEI formation lies in a compact and thin inorganic-rich SEI layer formed in the LiDFTFSI-based one, mainly attributed to the facile and exhaustive electro- or chemical reduction of DFTFSI– as a consequence of the replacement of CF3 with the CF2H group in TFSI–, as noted above in Figures 2D, 2E, and S2. The Chemical Stability of the Designer Anion with Polysulfide Since the high-order PS intermediates (Li2Sx, x = 3–8) formed during the discharge and charge process are soluble in PEO-based SPEs, the resistance of salt anions toward those aggressive PS species is of vital importance for the long-term cyclability of SSLSBs.30,32,39 Several Li salts identified in previous works are proved to be unsuited for SSLSBs, despite the notably enhanced stability toward Li anode in other battery systems. For example, Li bis(fluorosulfonyl)imide {Li[N(SO2F)2], LiFSI} was proposed as a potential replacement of LiTFSI in view of its good compatibility
1694 Joule 3, 1689–1702, July 17, 2019
Figure 4. The Chemical Stability of Designer Anion with Polysulfide (A) The appearance of 1 M LiX/DME (X = DFTFSI or TFSI) and blank DME solution before and after the addition of PS at room temperature for 60 h. (B) Normalized UV-vis absorption spectra of the PS-added solutions. (C) DFT calculations for the proposed intermediates. Light gray, red, yellow, light blue, dark blue, gray, and pink balls stand for H, O, S, F, N, C, and Li atoms, respectively.
with Li anode. However, recent works showed that LiFSI reacts spontaneously with PS species, which causes an irreversible consumption of electroactive materials, thus resulting in a decreased discharge capacity of SSLSBs.30,32,39 To this end, the chemical stabilities of both LiDFTFSI and LiTFSI versus PS were studied with a combination of experimental and theoretical approaches. Figure 4A shows the appearances of the LiX/DME (X = DFTFSI or TFSI) electrolytes before and after the addition of Li2S6 solution as well as the reference solution of Li2S6/DME for comparison. Upon addition of Li2S6, the color of both LiDFTFSI- and LiTFSI-based electrolytes
Joule 3, 1689–1702, July 17, 2019 1695
immediately changes from colorless to brownish, being close to that of the reference Li2S6/DME solution, and the color remains unchanged after 60 h of stirring at room temperature (Figure 4A). The negligible difference in the appearance is confirmed by the ultraviolet-visible (UV-vis) measurement, where nearly superimposed traces of those three solutions in the visible light region (390–700 nm) are observed (Figure 4B). These results suggest the good chemical stability of both LiDFTFSI and LiTFSI toward PS species. Considering the possible chemical lability of H in the CF2H group, three types of reactions between PS and DFTFSI– are conceived and assessed with density functional theory (DFT) calculations. The corresponding Cartesian coordinates of the optimized structures are summarized in Table S3. As shown in Figure 4C, the abstraction of the H atom from the CF2H group with Li2S6 is the most unlikely reaction pathway (path 1) because of the lowest stabilization energy (DE) compared to the other two reactions (96 kJ mol1 [path 1] versus 178 kJ mol1 [path 2] versus 233 kJ mol1 [path 3]). In particular, the cleavage of the C–S bond from the CF3 side in DFTFSI– delivers a similar energy to that in DFTFSI–, being much lower than the PS reactive anion, bis(fluorosulfonyl)imide anion (FSI–, DEFSI = 654 kJ mol1).32 This suggests that the nucleophilic reaction of both LiDFTFSI and LiTFSI with Li2S6 are not energetically favorable, which is consistent with the results from visual and UV-vis tests in Figures 4A and 4B. Electrochemical Performance of Designer Anion in SSLSB Inspired by the aforementioned superior physicochemical and electrochemical properties of LiDFTFSI versus LiTFSI, the Li-S cells using both PEO-based SPEs were assembled and cycled at 70 C. Figures 5A and 5B show the discharge and charge profiles at different C-rates. The LiDFTFSI-based cell delivers a higher specific discharge and areal capacity than the LiTFSI-based one, e.g., 1,035 mAh g1 and 1.08 mAh cm2 (LiDFTFSI) versus 666 mAh g1 and 0.77 mAh cm2 (LiTFSI) for the third cycle at a charge and discharge rate of 0.05/0.05C. This could be ascribed to the higher Li-ion conductivity of LiDFTFSI-based electrolytes,33 which facilitates a faster Li-ion diffusion and migration in both the composite S electrode and bulk electrolyte, thereby bringing a higher degree of S utilization. As shown in Figure 5C, the LiDFTFSI-based cell exhibits high discharge capacity and cycling stability (i.e., CE close to 100%), whereas the LiTFSI-based one encounters PS shuttling after a few cycles as noted by the significantly declining CE upon cycling (e.g., CE = 74% for the 15th cycle at 0.2C and endless charging at the 34th cycle shown in Figure S7). The electrochemical impedance (EIS) spectra measured under fully charged state (Figure S8) show a sharper slope at the low-frequency region for the cycled LiTFSIbased cell than that for the LiDFTFSI-based one, further implying the uncompleted conversion of low-order PS (e.g., Li2S, Li2S2) species to elemental sulfur in the former cell.45 These results clearly suggest that the SEI layer on Li anode formed in LiDFTFSI-based electrolyte is not only dense and compact, which allows a highly reversible and dendrite-free Li plating and stripping, but also resistant enough toward the nucleophilic attack of the soluble PS intermediates. This corroborates well with the above mechanistic understanding of the electrochemical behavior of Li electrode in LiDFTFSI-based electrolyte (Figures 2 and 3). The benefit is such that the LiDFTFSI-based cell sustains more than 1,300 cycles (3,000 h continuous cycling) with CEs close to 100% after imposing the C-rate to a discharge and charge rate of 0.1/0.1C (Figure 5D), in spite of the reduced discharge and charge capacity due to increased cell polarization (Figure S9). Please note that no attempt was made to optimize the sulfur electrode (ratios, mixing) with the LiDFTFSI electrolyte. In effect, the gravimetric energy density and cyclability of the LiDFTFSI-based cell
1696 Joule 3, 1689–1702, July 17, 2019
Joule 3, 1689–1702, July 17, 2019 1697
Figure 5. Electrochemical Performance of Designer Anion in SSLSB (A and B) Discharge/charge profiles of the Li-S cells using LiX/PEO (X = DFTFSI [A] or TFSI [B]) electrolytes at 70 C. (C) Rate capability of the Li-S cells with LiX/PEO (X = DFTFSI or TFSI) electrolytes. (D) Long-term cycling stability of the Li-S cells using LiDFTFSI/PEO electrolytes at a discharge and charge rate of 0.1/0.1C (after C-rate test). (E) Comparison of the cycling performance for SSLSBs reported in literature and the results of this work. The numbers in squares correspond to the entries listed in Table S4, and color code corresponds to the type of electrolyte. The calculation details for the gravimetric energy density are available in Supplemental Information.
outperforms all reported results from literature, as summarized in Figure 5E. Those results suggest the structural design of Li salt offers a fertile ground for regulating the electrochemical performance of SSLSBs, as the anion chemistry governs (1) the ionic transport behavior in both bulk electrolyte and S cathode and (2) the composition and morphology of the SEI layer on Li anode. As seen in Figure S10, with a moderate S loading of ca. 1.0 mg cm2 , the LiDFTFSIbased cells show the best cell performance in terms of areal capacity (i.e., gravimetric energy density). Further increase of S loading leads to a significantly lower S utilization and a pronounced PS shuttling, thereby decreasing the energy efficiency and density at cell level (e.g., 1,035 mAh g3 [1.11 mgsulfur cm2] versus 502 mAh g1 [1.59 mgsulfur cm2]). Though the capacity retention and attainable energy density of the LiDFTFSI-based cells currently fall behind the stringent requirements of practical Li-S batteries, we believe that further optimization of electrolyte formation and precise control of cathode architecture could sufficiently improve the performance of the LiDFTFSI-based SSLSBs. For example, the replacement of PEO with a fully amorphous polymer matrix, such as Jeffamine-based polymers,38,46,47 could improve the ionic conductivity of SPEs in the room temperature region; the implementation of nano-structural design of S/C composite and cathode additives, which could increase the S loading and utilization and minimize the shuttling effect of PS intermediates, thereby improving the energy density of SSLSBs.48–50
DISCUSSION The designer anion, DFTFSI, with the substitution of CF3 moiety in the widely used TFSI with CF2H moiety, is proposed for coping with the sophisticated cell chemistry of Li-S batteries. In addition to the enhanced Li-ion conductivity due to the selective introduction of hydrogen atom in sulfonimide anion, LiDFTFSI shows several unprecedented features that are essential for high-performance SSLSBs: (1) the asymmetric structure of DFTFSI– and the presence of electrochemically labile CF2H moiety induce a facile decomposition of DFTFSI– on Li anode, leading to the formation of several critical SEI building species, such as ionic conductive LiH and mechanically stable LiF; (2) though a hydrogen atom is presented in the anionic structure, DFTFSI– is still chemically stable against the aggressive PS intermediates formed during the cycling of Li-S batteries, thus ensuring the good chemical compatibility of LiDFTFSI with the cell chemistry of Li-S batteries; (3) among all the reported results of SSLSBs (Table S4), the LiDFTFSI-based Li-S cells show ultrahigh gravimetric energy density and remarkable capacity retention upon prolonged cycling because of the enhanced Li-ion conductivity, high-quality SEI formed on Li anode, and good chemical stability of DFTFSI– versus PS intermediates. Further improvements of the electrolyte formation and cathode architectural design would sufficiently enhance the energy density of LiDFTFSI-based SSLSBs and thereby make such technology better than the state-of-the-art Li-S batteries employing liquid electrolytes. Together with the easy accessibility of LiDFTFSI in terms of synthesis, the LiDFTFSI-based electrolytes are promising candidates for improving
1698 Joule 3, 1689–1702, July 17, 2019
the electrochemical performances of SSLSBs, thus promoting their practical deployment in the future.
EXPERIMENTAL PROCEDURES Preparation of Polymer Electrolytes LiDFTFSI was synthesized according to our recent work.33,34 Polymer membranes with an average thickness of 50 mm were prepared by conventional solvent casting method, followed by hot-pressing (High Temperature Film Maker Controller, Specac).33,34 The polymer was dissolved into acetonitrile, and then the salts, either LiDFTFSI or LiTFSI (battery grade, Solvionic, France), were added. In all cases, the salt concentration was 20 (the molar ratio of CH2CH2O [EO]/Li+). Electrochemical Stability of Electrolyte/Li Interphase Li symmetrical coin cells (Li jj Li ) were assembled in an argon-filled glove box to investigate the electrochemical stability between electrolyte and Li interphase. The galvanostatic cycling of the Li symmetric cells was evaluated using a VMP 3 potentiostat (Bio-Logic Science Instruments). The Li symmetric cells were cycled galvanostatically at a current density of 0.1 mA cm2, wherein the duration of each half-cycle was 2 h. Surface Morphology and Composition of the Cycled Li Electrode Surface morphologies of the Li deposits were examined by a field emission Quanta 200 FEG (FEI) microscope, operated at 20 kV. The Li deposits were obtained by the galvanostatic deposition of Li on Cu substrates using Li j LiX/DME (X = DFTFSI or TFSI) j Cu cells at a current density of 0.1 mA cm2 for 20 h. The compositions of the surface layer were measured by a Phoibos 150 XPS with a non-monochromatic Mg Ka source (hn = 1,253.6 eV). The spectra were recorded with high-resolution scans at low power (100 W, 20 eV pass energy, and 0.1 eV energy step). The calibration of the binding energy was performed taking into account as reference the Auger parameter of LiF at 1,340 eV.51 The samples were gently rinsed with DME and dried thoroughly under vacuum before transferred to the SEM or XPS chamber by a homedesigned airtight setup. Electrochemical Reduction Simulation Naphthalene was dissolved in ultra-pure and dry THF for about 30 min. Following, ground Li was added to the naphthalene and THF solution in a 1:1 mole ratio (naphthalene: Li). The reaction mixture was stirred for about 4 h at room temperature using a special glass-coated magnetic stirring bar as polytetrafluoroethylene (PTFE) is reduced at these potentials. Afterward, a dark green color, characteristic of the radical anion was observed (Figure S2). Following the complete formation of the radical anion, the reduction of both LiDFTFSI and LiTFSI was tested by adding the corresponding amounts of the salt. Finally, the residual color was monitored. Reaction of Polysulfide with Anions The solution of 0.1 M Li2S6/DME had been prepared by mixing Li2S and S with a molar ratio of 1:5 in DME, followed by stirring for 1 week at room temperature.42 The possible reaction between PS species and the two investigated anions were carried out in DME solution at room temperature. To a stirred solution of 1 M LiX/DME (X = DFTFSI or TFSI), a predetermined amount of 0.1 M Li2S6/DME solution with a molar ratio of LiX/Li2S6 of 200 was added. The UV-vis measurements were carried out, without dilution, on a Cary 5000 UV-vis spectrophotometer (Varian).
Joule 3, 1689–1702, July 17, 2019 1699
Our DFT calculations were based on the Becke’s three parameters (B3) exchange functional along with the Lee-Yang-Parr (LYP) non-local correlation functional (B3LYP),52,53 as implemented in the numeric atom-centered basis set all-electron code FHI-aims.54,55 The starting geometries of the investigated sulfonamide anions, considering different conformers, were generated by varying either S‒N‒S‒R1 or S‒N‒S‒R2 dihedral angles. The initial geometries were constructed using the open-source molecular editor and visualizer Avogadro.56 We used the universal force field and genetic algorithm search tool as implemented in Avogadro to prescreen dozens of low-energy geometries. Then, their atomic structures using a trust-radius-method-enhanced version of the Broyden-Fletcher-Goldfarb-Shanno optimization algorithm57 were fully relaxed. We used ‘‘tight’’ settings and ‘‘tier2’’ standard basis sets for H, Li, C, N, O, F, and S atoms. A maximum residual force component per atom of 0.01 eV A˚1 was the threshold for the convergence criteria in all structural relaxations, together with 104 electrons for the electron density and 106 eV for the total energy of the system. S Cathode Preparation and Cycling of Li-S Polymer Cell Composite sulfur cathode was prepared with 40 wt % elemental sulfur (99.5 wt %, Sigma-Aldrich), 15 wt % conductive carbons (Ketjen Black, KJ600, Akzo-Nobel), and 45 wt % LiX/PEO (X = DFTFSI or TFSI) as electrolyte. The S loading was from 1.0 to 1.1 mg cm2. The procedure has been detailed in our previous work.30,31 Li-S polymer cells were assembled in an argon-filled glovebox using the prepared electrode as cathode, polymer electrolyte membrane as both electrolyte and separator, and Li disk (China Energy Lithium) as anode. The cells were then cycled galvanostatically at a constant current mode between 1.6 and 2.8 V at 70 C using a BT lab cycler (Bio-Logic Science Instruments). Meanwhile, the EIS spectra of the cells in the fully charged state after various cycles were collected over the frequencies from 104 to 102 Hz using the above BT lab cycler.
SUPPLEMENTAL INFORMATION Supplemental Information can be found online at https://doi.org/10.1016/j.joule. 2019.05.003.
ACKNOWLEDGMENTS This work was supported by the Ministerio de Economı´a y Competitividad (MINECO) of the Spanish Government through Proyectos I + D Retos program (ENE201564907C2-1-R and ENE2016-81020-R grants). We also acknowledge funding by the Basque Government through the GVELKARTEK-2016 program. X.J. thanks the Basque Government for PhD funding, C.L. thanks the Spanish Government for the Juan de la Cierva scholarship (ref: FJCI-2015-23898), and H.Z. thank the Basque Government for the Berrikertu program (1-AFW-2017-2). We are also grateful for computer resources to the i2BASQUE academic network and SGI/IZOSGIker UPV/ EHU (Arina cluster) and the XPS measurements and discussions with Dr. Miguel A´ngel Mun˜oz-Ma´rquez. We thank Solvay, especially Claude Mercier, for the generous supply of difluoromethanesulfonyl chloride.
AUTHOR CONTRIBUTIONS H.Z. and M.A. conceived the research and designed the experiments. U.O., X.J., G.G.E., M.M.-I., and C.L. carried out the experiments and measurements. H.Z. and J.C. performed the computational work. All authors discussed the analysis of results. H.Z. and M.A. wrote the manuscript with input from all authors. H.Z. and M.A. supervised the work.
1700 Joule 3, 1689–1702, July 17, 2019
DECLARATION OF INTERESTS The authors declare no competing interests. Received: March 6, 2019 Revised: April 19, 2019 Accepted: May 3, 2019 Published: May 24, 2019
REFERENCES 1. Smalley, R.E. (2005). Future Global Energy Prosperity: The Terawatt Challenge. MRS Bulletin 30, 412–417.
16. Cao, R., Xu, W., Lv, D., Xiao, J., and Zhang, J.-G. (2015). Anodes for rechargeable lithium-sulfur batteries. Adv. Energy Mater 5, 1402273.
2. Dunn, B., Kamath, H., and Tarascon, J.M. (2011). Electrical energy storage for the grid: a battery of choices. Science 334, 928–935.
17. Lin, D., Liu, Y., and Cui, Y. (2017). Reviving the lithium metal anode for high-energy batteries. Nat. Nanotechnol 12, 194–206.
3. Tarascon, J.M., and Armand, M. (2001). Issues and challenges facing rechargeable lithium batteries. Nature 414, 359–367.
18. Cheng, X.B., Zhang, R., Zhao, C.Z., and Zhang, Q. (2017). Toward safe lithium metal anode in rechargeable batteries: a review. Chem. Rev. 117, 10403–10473.
4. Armand, M., and Tarascon, J.M. (2008). Building better batteries. Nature 451, 652–657. 5. Nitta, N., Wu, F., Lee, J.T., and Yushin, G. (2015). Li-ion battery materials: present and future. Mater. Today 18, 252–264. 6. Goodenough, J.B. (2018). How we made the Liion rechargeable battery. Nat. Electron 1, 204. 7. Cano, Z.P., Banham, D., Ye, S., Hintennach, A., Lu, J., Fowler, M., and Chen, Z. (2018). Batteries and fuel cells for emerging electric vehicle markets. Nat. Energy 3, 279–289. 8. Bruce, P.G., Freunberger, S.A., Hardwick, L.J., and Tarascon, J.M. (2011). Li-O2 and Li-S batteries with high energy storage. Nat. Mater 11, 19–29. 9. Choi, J.W., and Aurbach, D. (2016). Promise and reality of post-lithium-ion batteries with high energy densities. Nat. Rev. Mater 1, 16013. 10. Zhang, H., Eshetu, G.G., Judez, X., Li, C., Rodriguez-Martı´nez, L.M., and Armand, M. (2018). Electrolyte additives for lithium metal anodes and rechargeable lithium metal batteries: progresses and perspectives. Angew. Chem. Int. Ed. 57, 15002–15027. 11. Shen, Y., Zhang, Y., Han, S., Wang, J., Peng, Z., and Chen, L. (2018). Unlocking the energy capabilities of lithium metal electrode with solid-state electrolytes. Joule 2, 1674–1689. 12. Judez, X., Eshetu, G.G., Li, C., RodriguezMartinez, L.M., Zhang, H., and Armand, M. (2018). Opportunities for rechargeable solidstate batteries based on Li-intercalation cathodes. Joule 2, 2208–2224. 13. Yin, Y.X., Xin, S., Guo, Y.G., and Wan, L.J. (2013). Lithium-sulfur batteries: Electrochemistry, materials, and prospects. Angew. Chem. Int. Ed. Engl. 52, 13186–13200. 14. Nazar, L.F., Cuisinier, M., and Pang, Q. (2014). Lithium-sulfur batteries. MRS Bull. 39, 436–442. 15. Xu, W., Wang, J., Ding, F., Chen, X., Nasybulin, E., Zhang, Y., and Zhang, J.-G. (2014). Lithium metal anodes for rechargeable batteries. Energy Environ. Sci. 7, 513–537.
19. Fenton, D.E., Parker, J.M., and Wright, P.V. (1973). Complexes of alkali metal ions with poly(ethylene oxide). Polymer 14, 589. 20. Armand, M., Chabagno, J.M., and Duclot, M.J. (1978). Second International Meeting on Solid Electrolytes Extended Abstracts. St. Andrews Ecosse. September 20–22. 21. Armand, M.B. (1986). Polymer electrolytes. Annu. Rev. Mater. Sci. 16, 245–261. 22. Di Noto, V., Lavina, S., Giffin, G.A., Negro, E., and Scrosati, B. (2011). Polymer electrolytes: present, past and future. Electrochim. Acta 57, 4–13. 23. Hallinan, D.T., and Balsara, N.P. (2013). Polymer electrolytes. Annu. Rev. Mater. Res. 43, 503–525. 24. Zhang, H., Li, C., Piszcz, M., Coya, E., Rojo, T., Rodriguez-Martinez, L.M., Armand, M., and Zhou, Z. (2017). Single lithium-ion conducting solid polymer electrolytes: advances and perspectives. Chem. Soc. Rev. 46, 797–815. 25. Li, C., Zhang, H., Otaegui, L., Singh, G., Armand, M., and Rodriguez-Martinez, L.M. (2016). Estimation of energy density of Li-S batteries with liquid and solid electrolytes. J. Power Sources 326, 1–5. 26. Judez, X., Zhang, H., Li, C., Eshetu, G.G., Gonza´lez-Marcos, J.A., Armand, M., and Rodriguez-Martinez, L.M. (2018). Review— Solid electrolytes for safe and high energy density lithium-sulfur batteries: promises and challenges. J. Electrochem. Soc 165, A6008– A6016. 27. Xu, K. (2014). Electrolytes and interphases in Liion batteries and beyond. Chem. Rev. 114, 11503–11618. 28. Younesi, R., Veith, G.M., Johansson, P., Edstro¨m, K., and Vegge, T. (2015). Lithium salts for advanced lithium batteries: Li–metal, Li–O2, and Li–S. Energy Environ. Sci. 8, 1905–1922. 29. Zhang, H., Liu, C., Zheng, L., Xu, F., Feng, W., Li, H., Huang, X., Armand, M., Nie, J., and Zhou, Z. (2014). Lithium bis(fluorosulfonyl)imide/
poly(ethylene oxide) polymer electrolyte. Electrochim. Acta 133, 529–538. 30. Judez, X., Zhang, H., Li, C., Gonza´lez-Marcos, J.A., Zhou, Z., Armand, M., and RodriguezMartinez, L.M. (2017). Lithium bis(fluorosulfonyl) imide/poly(ethylene oxide) polymer electrolyte for all solid-state Li–S cell. J. Phys. Chem. Lett. 8, 1956–1960. 31. Eshetu, G.G., Judez, X., Li, C., Bondarchuk, O., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2017). Lithium azide as an electrolyte additive for all-solid-state lithium– sulfur batteries. Angew. Chem. Int. Ed. Engl. 56, 15368–15372. 32. Eshetu, G.G., Judez, X., Li, C., Martinez-Iban˜ez, M., Gracia, I., Bondarchuk, O., Carrasco, J., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2018). Ultrahigh performance all solid-state lithium sulfur batteries: salt anion’s chemistry-induced anomalous synergistic effect. J. Am. Chem. Soc. 140, 9921–9933. 33. Zhang, H., Oteo, U., Zhu, H., Judez, X., Martinez-Iban˜ez, M., Aldalur, I., Sanchez-Diez, E., Li, C., Carrasco, J., Forsyth, M., et al. (2019). Enhanced lithium-ion conductivity of polymer electrolytes by selective introduction of hydrogen into the anion. Angew. Chem. Int. Ed. Engl. https://doi.org/10.1002/ange. 201813700. 34. Oteo, U., Martinez-Iban˜ez, M., Aldalur, I., Sanchez-Diez, E., Carrasco, J., Armand, M., and Zhang, H. (2019). Improvement of the cationic transport in polymer electrolytes with (difluoromethanesulfonyl) (trifluoromethanesulfonyl)imide salts. ChemElectroChem 6, 1019–1022. 35. Sun, X.-G., and Kerr, J.B. (2006). Synthesis and characterization of network single ion conductors based on comb-branched polyepoxide ethers and lithium bis(allylmalonato)borate. Macromolecules 39, 362–372. 36. Ikeya, M. (1977). Electrical properties of lithium hydride. J. Phys. Soc. Jpn 42, 168–174. 37. Lin, Z., Liu, Z., Dudney, N.J., and Liang, C. (2013). Lithium superionic sulfide cathode for all-solid lithium–sulfur batteries. ACS Nano 7, 2829–2833. 38. Aldalur, I., Martinez-Iban˜ez, M., Piszcz, M., Zhang, H., and Armand, M. (2018). Selfstanding highly conductive solid electrolytes based on block copolymers for rechargeable all-solid-state lithium-metal batteries. Batteries & Supercaps 1, 149–159. 39. Judez, X., Zhang, H., Li, C., Eshetu, G.G., Zhang, Y., Gonza´lez-Marcos, J.A., Armand, M.,
Joule 3, 1689–1702, July 17, 2019 1701
and Rodriguez-Martinez, L.M. (2017). Polymer-rich composite electrolytes for all-solid-state Li–S cells. J. Phys. Chem. Lett. 8, 3473–3477. 40. Qian, J., Henderson, W.A., Xu, W., Bhattacharya, P., Engelhard, M., Borodin, O., and Zhang, J.G. (2015). High rate and stable cycling of lithium metal anode. Nat. Commun 6, 6362. 41. Aurbach, D., Daroux, M.L., Faguy, P.W., and Yeager, E. (1988). Identification of surface films formed on lithium in dimethoxyethane and tetrahydrofuran solutions. J. Electrochem. Soc. 135, 1863–1871. 42. Suo, L., Hu, Y.S., Li, H., Armand, M., and Chen, L. (2013). A new class of solvent-in-salt electrolyte for high-energy rechargeable metallic lithium batteries. Nat. Commun 4, 1481. 43. Eshetu, G.G., Diemant, T., Grugeon, S., Behm, R.J., Laruelle, S., Armand, M., and Passerini, S. (2016). In-depth interfacial chemistry and reactivity focused investigation of lithium– imide- and lithium–imidazole-based electrolytes. ACS Appl. Mater. Interfaces 8, 16087–16100. 44. Aurbach, D., and Weissman, I. (1999). On the possibility of LiH formation on Li surfaces in wet electrolyte solutions. Electrochem. Commun 1, 324–331. 45. Deng, Z., Zhang, Z., Lai, Y., Liu, J., Li, J., and Liu, Y. (2013). Electrochemical impedance
1702 Joule 3, 1689–1702, July 17, 2019
spectroscopy study of a lithium/sulfur battery: modeling and analysis of capacity fading. J. Electrochem. Soc. 160, A553–A558. 46. Aldalur, I., Martinez-Iban˜ez, M., Piszcz, M., Rodriguez-Martinez, L.M., Zhang, H., and Armand, M. (2018). Lowering the operational temperature of all-solid-state lithium polymer cell with highly conductive and interfacially robust solid polymer electrolytes. J. Power Sources 383, 144–149. 47. Aldalur, I., Martinez-Iban˜ez, M., Krzto nMaziopa, A., Piszcz, M., Armand, M., and Zhang, H. (2019). Flowable polymer electrolytes for lithium metal batteries. J. Power Sources 423, 218–226. 48. Ma, J., Fang, Z., Yan, Y., Yang, Z., Gu, L., Hu, Y.-S., Li, H., Wang, Z., and Huang, X. (2015). Novel large-scale synthesis of a C/S nanocomposite with mixed conducting networks through a spray drying approach for Li–S batteries. Adv. Energy Mater 5, 1500046. 49. Yang, Y., Zheng, G., and Cui, Y. (2013). Nanostructured sulfur cathodes. Chem. Soc. Rev. 42, 3018–3032. 50. Pang, Q., Shyamsunder, A., Narayanan, B., Kwok, C.Y., Curtiss, L.A., and Nazar, L.F. (2018). Tuning the electrolyte network structure to invoke quasi-solid state sulfur conversion and suppress lithium dendrite formation in Li–S batteries. Nat. Energy 3, 783–791.
51. Wagner, C.D., Riggs, W.M., Davis, L.E., Moulder, J.F., and Muilenberg, G.E. (1979). A Reference Book of Standard Data for Use in X-Ray Photoelectron Spectroscopy (PerkinElemer). 52. Lee, C., Yang, W., and Parr, R.G. (1988). Development of the Colle-Salvetti correlationenergy formula into a functional of the electron density. Phys. Rev. B Condens. Matter 37, 785–789. 53. Becke, A.D. (1993). Density-functional thermochemistry. III. The role of exact exchange. J. Chem. Phys. 98, 5648–5652. 54. Blum, V., Gehrke, R., Hanke, F., Havu, P., Havu, V., Ren, X., Reuter, K., and Scheffler, M. (2009). Ab initio molecular simulations with numeric atom-centered orbitals. Comput. Phys. Commun 180, 2175–2196. 55. Havu, V., Blum, V., Havu, P., and Scheffler, M. (2009). Efficient integration for all-electron electronic structure calculation using numeric basis functions. J. Comput. Phys. 228, 8367– 8379. 56. Hanwell, M.D., Curtis, D.E., Lonie, D.C., Vandermeersch, T., Zurek, E., and Hutchison, G.R. (2012). Avogadro: an advanced semantic chemical editor, visualization, and analysis platform. J. ChemInform 4, 17. 57. Nocedal, J., and Wright, S.J. (2006). Sequential Quadratic Programming (Springer).