Stability of Glyme Solvate Ionic Liquid as an Electrolyte for

Jan 25, 2017 - The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acsami.6b14449. ...... Hardy , J...
1 downloads 14 Views 1MB Size
Subscriber access provided by University of Newcastle, Australia

Article

Stability of Glyme Solvate Ionic Liquid as Electrolyte for Rechargeable Li-O2 Battery Hoi-Min Kwon, Morgan L. Thomas, Ryoichi Tatara, Yoshiki Oda, Yuki Kobayashi, Azusa Nakanishi, Kazuhide Ueno, Kaoru Dokko, and Masayoshi Watanabe ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.6b14449 • Publication Date (Web): 25 Jan 2017 Downloaded from http://pubs.acs.org on January 26, 2017

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

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

Page 1 of 30

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

ACS Applied Materials & Interfaces

Stability of Glyme Solvate Ionic Liquid as Electrolyte for Rechargeable Li/O2 Battery Hoi-Min Kwon,† Morgan L. Thomas,† Ryoichi Tatara, Yoshiki Oda, Yuki Kobayashi, Azusa Nakanishi, Kazuhide Ueno, Kaoru Dokko, and Masayoshi Watanabe* Department of Chemistry and Biotechnology, Yokohama National University, 79-5 Tokiwadai, Hodogaya-ku, Yokohama 240-8501, Japan

ABSTRACT: A solvate ionic liquid (SIL) was compared with a conventional organic solvent for the electrolyte of the Li-O2 battery. An equimolar mixture of triglyme (G3) and lithium bis(trifluoromethanesulfonyl)amide (Li[TFSA]), and a G3/Li[TFSA] mixture containing excess glyme were chosen as the SIL and the conventional electrolyte, respectively. Charge behavior and accompanying gas evolution of the two electrolytes was investigated by electrochemical mass spectrometry (ECMS). From the linear sweep voltammetry performed on an as-prepared cell, we demonstrate that the SIL has a higher oxidative stability than the conventional electrolyte, and further offers the advantage of lower volatility, which would benefit an opentype lithium-O2 cell design. Moreover, CO2 evolution during galvanostatic charge was less in the SIL, which implies less side reaction. However, O2 evolution during charge did not reach the theoretical value in either of the two electrolytes. Several mass spectral fragments were generated during the charge process, which provided evidences for side reactions of glyme-based 1 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

electrolytes. We further relate the difference in observed discharge product morphology for these electrolytes to the solubility of the superoxide intermediate, determined by rotating ring disk electrode (RRDE) measurements.

KEYWORDS: lithium-oxygen battery, solvate ionic liquid, glyme, electrochemical mass spectrometry, rotating ring disk electrode voltammetry

1. INTRODUCTION The non-aqueous lithium-oxygen (Li-O2) battery has recently attracted great attention, in part due to the higher theoretical energy density than that of conventional lithium-ion batteries.1 However, there are several obstacles to the practical application of the Li-O2 cell. Those include volatility of the electrolytes, high charge overpotential, and the high reactivity of the reaction intermediate (the superoxide radical anion). Indeed, many conventional electrolytes such as propylene carbonate are now known to be unsuitable for use in the Li-O2 battery, because these electrolytes react with the discharge products.2–4 Longer chain glymes (glycol ethers, typically the dimethyl ethers of ethylene glycol oligomers) have been widely promoted as promising electrolytes for the Li-O2 battery. Glymes, such as triglyme (triethylene glycol dimethyl ether) and tetraglyme (tetraethylene glycol dimethyl ether), have relatively low vapor pressure, also there are numerous reports on the quasi-reversible reduction and oxidation of O2 in glymes.3,5–7 From theoretical calculations, glymes were judged to be stable with regard to nucleophilic substitution by O2− radicals.8 Moreover, concerning the full Li-O2 cell, the main discharge product using glymes was Li2O2, unlike carbonate electrolytes.3,7 Additionally, glymes that have a relatively low highest occupied molecular orbital (HOMO) level and high pKa show higher stability because solvents having higher HOMO levels are prone to oxidation during charging 2 ACS Paragon Plus Environment

Page 2 of 30

Page 3 of 30

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

ACS Applied Materials & Interfaces

and low pKa solvents are unstable with respect to H-abstraction during discharging, respectively.9–11 Indeed, the afore-mentioned characteristics indicate glymes are promising candidates for the Li-O2 battery. Solvate ionic liquids (SILs) are a new class of ILs and are exemplified by the equimolar mixture of certain glyme solvents and lithium salts, which show properties typical of ionic liquids, such as negligible vapor pressure and high oxidative stability.12 Also, we previously reported that SILs improve performance for the lithium-sulfur battery resulting from the prevention of dissolution of lithium polysulfides.13,14 The prohibited dissolution of the polysulfides owes to low Lewis acidity and low Lewis basicity of the cations and anions of the SILs, respectively, which greatly lowers interaction with lithium polysulfides. We would also assume that this prevention of the intermediate dissolution could improve the rechargeability of a Li-O2 battery, because the discharge product should be formed directly on the cathode, with little loss of the intermediate (O2−•) into solution, for a better reversibility. The properties of SILs described above led us to study a SIL electrolyte for Li-O2 batteries because it satisfies requirements such as high oxidative stability, negligible volatility, and stability towards lithium metal. Previously, we reported that quasi-reversible ORR/OER (oxygen reduction reaction / oxygen evolution reaction) occurs in [Li(G3)1][TFSA].6 Adams et al. employed a similar “chelate ionic liquid” based on a alkyl-substituted glyme, 2,3-dimethyl-2,3-dimethoxybutane (DMDMB), which showed improved stability in the Li-O2 system.15 A recent report has suggested that a highly concentrated Na+ electrolyte may exhibit improved stability in the presence of sodium superoxide, suggesting that the SIL employed in our study may also show improved stability.16 The remaining issue of high charge overpotential will not be addressed in

3 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

this work, but existing efforts on, for example, redox mediators may allow improvements in future.17 In this work, we expand upon our earlier investigation of the use of a solvate ionic liquid as electrolyte for the Li-O2 battery, and in particular we study the possibility for enhanced performance as compared to a conventional electrolyte solution. Specifically, [Li(G3)1][TFSA] and [Li(G3)4][TFSA] are used as the SIL and a representative conventional organic electrolyte, respectively. They are compared through electrochemical mass spectrometry (ECMS), which is chosen because it provides simultaneous qualitative and quantitative analysis of evolution of charge / discharge product (gas phase) and electrochemical measurements. Several other groups have previously used this method in the study of Li-O2 batteries.3,18–26 From the ECMS results we obtained further support for the possibility of using [Li(G3)1][TFSA] as a stable electrolyte in Li-O2 cells.

2. EXPREIMENTAL SECTION 2.1. Materials Triglyme (G3, with water content below 50 ppm) were obtained from Nippon Nyukazai Co., Ltd., Japan and used as received. Lithium bis(trifluoromethanesulfonyl)amide (Li[TFSA]) was kindly supplied by Solvay, Japan. [Li(G3)1][TFSA] and [Li(G3)4][TFSA] were prepared by mixing G3 and Li[TFSA] in a 1:1 molar ratio and in a 4:1 molar ratio, respectively. Ketjen black (KB, Lion Corp., Japan) was used for the O2 electrode. A slurry was made by mixing KB and PTFE binder (Sigma-Aldrich) in N-methyl-2-pyrrolidinone (Wako, Japan). O2 electrodes were prepared by coating the slurry onto carbon paper (TGP-H-060, Toray, Japan). Each electrode was punched into a 16 mm diameter circle. The carbon loading of the carbon electrodes

4 ACS Paragon Plus Environment

Page 4 of 30

Page 5 of 30

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

ACS Applied Materials & Interfaces

(excluding carbon paper) was about 0.7 mg/cm2. O2, He and mixed gases (He gas containing a known concentration of O2 or CO2) for the calibration encapsulated in steel containers (Suzuki Shokan, Japan) were used as received. The electrolytes were prepared by mixing solvents and Li salts in an Ar-filled glove box (VAC, [H2O]~0.5 ppm).

2.2 Measurements An ECC-Air electrochemical test cell (EL-Cell, Germany) was used for electrochemical measurements. The test cell was assembled in a glovebox under an Ar atmosphere, and comprised the O2 electrode described above, a porous separator (GA-55, Advantec, Japan), a Li metal foil (Honjo Metal, Japan) anode and the electrolyte (80 µL). Before the battery discharge test, the atmosphere of each test cell were substituted to O2. Electrochemical measurements were conducted with a Biologic VMP-3. For linear sweep voltammetry (LSV), the scan rate was 0.1 mV/s. Galvanostatic discharge-charge tests were performed at 0.1 mA/cm2. For MS, the EL-Cell was connected to a GCMS-QP2010 Ultra (Shimadzu, Japan), allowing electrochemical and MS measurements (quadrupole detector) to be conducted simultaneously. For the measurement of discharged cell, the cell was attached with by swift connection to minimalize the exposure of the interior contents. He (99.999 vol%) was used for the carrier gas. The flow rate of the carrier gas and the split ratio were 100 mL min-1 and 42.74:1, respectively. For all ECMS intensity data reported in this study, the signal was normalized by the measured intensity at m/z = 219 for a standard sample of perfluorotributylamine before each experiment (i.e. auto-tuning), allowing numerical comparison

5 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

of the intensity signals for different experiments. Additionally, for m/z = 32 and 44 (assigned to O2 and CO2), calibration was conducted to allow fully quantitative measurements to be performed. Calibrations were conducted using standard gases. Mixed gases flowed into the MS through a separated line. O2 and CO2 concentration was controlled by varying the ratio of mixed gas, and pure He and MS intensities of each concentration were recorded. Based on this relation, calibration curves were generated. The calibration curves are shown in Figure S1.All measurements were conducted at 30 °C. The measurements during charge were conducted after discharging up to 1 mAh. The electrochemical properties of ORR/OER in the electrolytes were investigated by cyclic voltammetry (CV) using rotating ring disk electrode (RRDE) in a water jacket type air-tight three electrode cell. The cell consisted of glassy carbon (GC) disk / GC ring electrode (disk electrode: d = 4 mm, ring electrode: dinside = 5 mm, doutside = 7 mm) as the working electrode, Li metal immersed in 1 mol dm−3 Li[TFSA]/G3 as the reference electrode and a Pt coil as the counter electrode. The reference electrode was separated from the working electrolyte using a liquid junction and kept in an Ar atmosphere. CVs using RRDE were recorded using an ALS 700E bipotentiostat, and collected at 1000 rpm and 100 m Vs−1 while the GC ring electrode was biased to 3.3 V. Temperature was maintained by using a water circulator at 30 ºC. The electrolytes were bubbled with dry O2 gas prior to use, and CV was conducted with continuous gas flow. FE-SEM (Field Emission - Scanning Electron Microscopy) images were measured on a Hitachi SU8010 scanning electron microscope at an accelerating voltage of 5 kV. An air-tight SEM specimen holder was used to prevent air exposure of the sample as follows. The sample was mounted on the specimen holder in an Ar-filled glovebox, and then sealed with an air-tight

6 ACS Paragon Plus Environment

Page 6 of 30

Page 7 of 30

cover. The cover was removed in the sample exchange (loading) chamber under vacuum, preventing exposure of the sample to air.

10

(a) [Li(G3)4][TFSA]

8

6

6 4 4 2

2

0

Relative abundance / %

8

Current / mA

Normalized MS intensity / a.u.

3. RESULTS AND DISCUSSION

0

100 75 50 25 0 100 75 50 25 0 100 75 50 25 0

(c)

4.00 V

4.75 V

5.00 V

0

20

40

60

80

100

m/z 10

(b) [Li(G3)1][TFSA]

8

6

6 4 4 2

2

0

0 3.0

3.5 4.0 4.5 + E / V vs. Li/Li

Relative abundance / %

8

Current / mA

Normalized MS intensity / a.u.

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

ACS Applied Materials & Interfaces

100 75 50 25 0 100 75 50 25 0 100 75 50 25 0

4.75 V

5.00 V

0

5.0

4.00 V

(d)

20

40

60

m/z

7 ACS Paragon Plus Environment

80

100

ACS Applied Materials & Interfaces

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

Figure 1. Current and intensity of m/z = 15 mass fragment during LSV of as prepared cells under helium at 0.1 mV s−1 with (a) [Li(G3)4] [TFSA] or (b) [Li(G3) 1][TFSA] electrolytes, and corresponding overall mass spectra (c, d) for the points 4.00 V, 4.75 V and 5.00 V. Y-axes in (a) and (b) show the same range for ease of comparison. Figures 1(a) and (b) show the current flow during LSV performed for the pristine electrodes with [Li(G3)4][TFSA] (glyme excess electrolyte) and [Li(G3)1][TFSA] (SIL) respectively. The two electrolyte systems showed notably different behaviors. In the lower potential area (< 4.5 V), no noticeable current flow was observed, while appreciable current flows were detected above 4.5 V. Similar large anodic currents were previously attributed to oxidative decomposition of glyme molecules.12 Both systems ([Li(G3)4][TFSA] and [Li(G3)1][TFSA]) exhibited currents in the high potential area (> 4.5 V) while the magnitude of current was different. [Li(G3)4][TFSA] exhibited a much higher current flow, which was 4.5 times greater at 5.0 V. In addition, the observation of significant signals of m/z = 15 in the high potential area supports the proposal that the observed current can be attributed to decomposition of glymes. The selection of m/z = 15 as indicative of glyme decomposition here is based on consideration of the overall mass spectra recorded during the LSV measurements, see Figure 1(c) and 1(d), as will be further described below. It should be noted that the signal of varying intensity at m/z = 18 in these figures is attributed to H2O and results from the instrumental background. In Figure 1(c), at a potential of 4.00 V vs. Li/Li+, there are clusters of mass fragment signals at m/z values of ~30, 45 and 60, with a similar ratio to the spectrum of pure G3 (see Figure S2). This is attributed to evaporation of excess glyme (i.e. free glyme), and indicates a cell constructed with [Li(G3)4][TFSA] electrolyte would undergo the gradual loss of glyme during operation. Thus, although G3 has a relatively low vapor pressure, it appears that solvent 8 ACS Paragon Plus Environment

Page 8 of 30

Page 9 of 30

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

ACS Applied Materials & Interfaces

evaporation is inevitable. As the potential is swept to more positive values, the ratio of the characteristic G3 fragments shifts to lower values of m/z, as shown in Figure 1(c) for 4.75 V and 5.00 V. This change in the ratio of the fragments is attributed to decomposition of G3, towards lower molecular weight components with mass spectra at lower values of m/z. This shift to lower m/z ratios can be observed in considering the series G3 > G2 > G1 (Figure S2). The m/z = 15 signal is particularly revealing. This mass fragment can reasonably be assumed to result only from a CH3 moiety (originating from the glyme rather than the anion, which contains no protons).27 It clearly increases in intensity relative to the most intense signal for pure G3 (m/z = 59), and thus is used here as an indication of partial oxidative decomposition of glyme into lower molecular weight fragments. This, and other mass fragments considered in this work are summarized in Table 1.

Table 1. Mass fragments considered in this work, with speculative assignments

m/z

Elemental formula

Assignment

15

CH3+

G3 decomposition products

32

O2+

Dioxygen (oxidation of Li2O2)

44

CO2+

Carbon Dioxide (Li2CO3 or LiRCO3)

69

CF3+

TFSA− decomposition products

73

C3H5O2+

G3 decomposition products

9 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

Page 10 of 30

In contrast, for [Li(G3)1][TFSA], the intensity of m/z = 15 remains quite low until a high potential is reached. Moreover, signals characteristic of pure G3 are not observed. This is because [Li(G3)1][TFSA] has low volatility, i.e. the thermal stability of the glyme is improved in the 1:1 complex.28 The current and corresponding mass signals indicate that [Li(G3)1][TFSA] is more stable at high potential. This observation arises from the difference in oxidative stabilities of the two electrolytes. As we previously reported, [Li(G3)1]+ has high oxidative stability due to the lowering of the HOMO energy levels upon complexation of the Li+ with glyme.12 In [Li(G3)1][TFSA], more than 97 % of glymes exist in complex form, i.e. [Li(G3)1]+, when measured by Raman,29 i.e. there is only a small amount of free glyme, and hence the oxidative stability of the 1:1 complex is higher than [Li(G3)4][TFSA] containing a large excess of free glyme. In many previous reports, a large overpotential for charge was observed and this induces

10 ACS Paragon Plus Environment

Page 11 of 30

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

ACS Applied Materials & Interfaces

the decomposition of the electrolyte solution.3,18,30–32 Therefore, for Li-O2 batteries, improving the oxidative stability of the electrolytes is important due to the large charge overpotential. In terms of the volatility and oxidative stability, [Li(G3)1][TFSA] is more suitable for Li-O2 batteries.

Figure 2. LSV of galvanostatically discharged cathodes and simultaneously measured m/z = 32 under He at 0.1 mV s−1 in (a) [Li(G3)4][TFSA] and (b) [Li(G3)1][TFSA]. Insets show m/z = 32, 44 signals and the current through the whole voltage range.

11 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

Page 12 of 30

Figure 2 shows the LSV results using cells previously discharged to 1 mAh (discharge curves are shown in Figure S3). They exhibit markedly different behavior compared to the pristine electrodes. In Figure 2, both systems exhibit oxidation current and O2 evolution (m/z = 32) having an onset at around 3 V. The change in the O2 signal intensity coincides with the current flow. These are assigned to the following reactions (either a combination of (1) and (2)33,34, or (3) directly35,36) at least at lower potentials than 4.25 V. Li2O2 → LiO2 + Li+ + e−

(1)

LiO2 + LiO2 → Li2O2 + O2

(2)

Li2O2 → 2Li+ + O2 + 2e−

(3)

At higher potentials, following the moderate current and O2 evolution, vigorous current and MS fragment signals (m/z = 32, 44) were observed, especially for [Li(G3)4][TFSA] (Figure 2 inset. See also selected mass spectra in Figure S4). O2 observed in this potential range can also be assigned to oxidation of Li2O2.37 The m/z = 44 signal can be assigned to CO2 evolution. The concurrent detection of m/z = 28 (CO+) and m/z = 16 (O+) signals supports this assignment (see Figure S5). Similar behavior in DME (dimethoxyethane) was reported by McCloskey et al.,38 and they suggested that CO2 evolution at high potential is from the oxidation of Li2CO3 (or LiRCO3). This evolution of CO2 also could be assigned to the oxidation of side products such as lithium formate and acetate.2 Compared to [Li(G3)4][TFSA], [Li(G3)1][TFSA] showed lower intensities (current and mass signals) for side product oxidation. From this, we can conclude that the quantity of side products such as Li2CO3 is greater in [Li(G3)4][TFSA] than in [Li(G3)1][TFSA].

12 ACS Paragon Plus Environment

Page 13 of 30

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

ACS Applied Materials & Interfaces

Figure 3. Electrochemical MS data for cells constructed with either [Li(G3)1][TFSA] (a, c) or [Li(G3)4][TFSA] (b, d) electrolytes after galvanostatic discharge to 1 mAh. From top, galvanostatic charge curves (a, b), corresponding O2 (blue), CO2 (red) evolution rates (c, d). The blue (c,d) and black (e) broken lines denotes theoretical O2 evolution. Integrated O2 evolution (e) and CO2 evolution (f) during charge, which was calculated from (c) and (d).

13 ACS Paragon Plus Environment

ACS Applied Materials & Interfaces

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

Page 14 of 30

In order to compare the difference between the electrolytes under more realistic cell conditions, ECMS was also measured during galvanostatic charging. In comparison of the charge curves, [Li(G3)1][TFSA] and [Li(G3)4][TFSA] showed further differences (Figure 3 (a, b)). In [Li(G3)4][TFSA], the initial charging overvoltage is large and then a plateau is observed close to 4.2 V. In [Li(G3)1][TFSA], the voltage rises over the entire charging process. This is similar to the trend for cumulative current measured by LSV (Figure S6). We assume that this difference of charge behavior is due to a difference of discharge product morphology. 39,40 This will be further

elaborated in the discussion below.

14 ACS Paragon Plus Environment

Page 15 of 30

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

ACS Applied Materials & Interfaces

Figure 4. FE-SEM images of discharged cathodes. The discharge capacity was 0.75 mAh cm−2 on KB electrode for (a), (b) and 0.0275 mAh cm−2 on the bare carbon paper for (c), (d). (a), (c) [Li(G3)4][TFSA] and (b), (d) [Li(G3)1][TFSA] were used as the electrolytes, respectively.

In the FE-SEM observation, we found the morphology of discharge products for these two systems is different (Figure 4 (a), (b)). The discharge product formed in [Li(G3)4][TFSA] exhibits a needle-like structure. Griffith et al reported this shape of Li2O2 that formed at high discharge current.41 We assume that this needle-like shape is formed at the initial stage of the discharge process and finally it may form a typical toroid-like morphology. In contrast, [Li(G3)1][TFSA] showed no needle or toroidal-like morphology, but exhibited a film-like discharge product. To further assess the difference in discharge product morphology, we used a carbon paper-only cathode (Figure 4 (c), (d)). Although the achievable discharge capacity for such electrodes is rather low, it is more simple to distinguish the discharge product from a cathode material consisting only of carbon fibers, as compared to a composite electrode. SEM images after discharge using only carbon paper as a cathode also showed a different morphology of the discharge products in the two electrolytes. It should be noted that this difference between the two electrolytes is similar to the difference caused by water content.42,43 However, both electrolytes had very low water content (