Sulfur‐containing compounds as electrolyte additives for lithium‐ion batteries

Abstract

Originating from “rocking‐chair concept”, lithium‐ion batteries (LIBs) have become one of the most important electrochemical energy storage technologies, which have largely impacted our daily life. The utilization of electrolyte additives in small quantities (≤5% by wt or vol) has been long viewed as an economical and efficient approach to regulate the properties of electrolyte and electrode–electrolyte interphases and consequently improve the cycling performance of LIBs. Among all the kinds of electrolyte additives, sulfur‐containing compounds have gained significant attention due to their unique features in building stable electrode–electrolyte interphases and protect battery cells from overcharging. In this work, advances and progresses of sulfur‐containing additives used in LIBs are overviewed, with special attention paid to the working mechanisms of these electrolyte additives. Particularly, four representative sulfur‐containing compounds (i.e., 1,3‐propane sultone, prop‐1‐ene‐1,3‐sultone, 1,3,2‐dioxathiolane‐2,2‐dioxide, and ethylene sulfite) are comparatively discussed concerning their impact on electrode–electrolyte interphases and cell performances. Future work on the development of sulfur‐containing compounds as functional electrolyte additives is also provided. The present review is anticipated to be not only a base document to access the status quo in this research domain but also a guideline to select specialized additives and electrolytes for practical applications.

INTRODUCTION

Lithium-ion batteries (LIBs) have been commercialized since the early 1990s¹ and are used widely in consumer electronic products, electric vehicles (EVs), and electric grids.^2–5 LIBs are also known as “rocking-chair batteries” as the battery cell is conceptually built on the reversible (de)intercalation of Li⁺ ions, during charge and discharge processes, into two electrode materials with different redox potentials.^6–8 Typical LIBs mainly consist of an intercalation-type negative (i.e., graphite, Li₄Ti₅O₁₂, etc.) and a positive electrode (i.e., LiCoO₂, LiNi_1-x-yCo_xMn_yO₂, LiFePO₄, LiNi_0.5Mn_1.5O₄, etc.), polymeric separators (e.g., polypropylene and polyethylene), and nonaqueous electrolytes.^2,3 Among these main components of LIBs, the Li-ion conductive electrolytes not only control the transportation of active species (i.e., Li⁺ ions) during battery cycling but also actively participate in the formation of electrode–electrolyte interphases. Therefore, the intrinsic characters of electrolyte materials have been viewed as one of the most essential parameters in dictating the cycling stability, calendar life, and storage performance of LIBs.^2,3

The nonaqueous electrolytes used in LIBs commonly comprise organic solvents, conductive lithium (Li) salts, and functional additives.^2,3 As schematically shown in Figure 1(A), the electrolytes are unstable toward the highly reducing negative electrode and the highly oxidizing positive electrode, and reductive/oxidative decomposition of electrolyte components on the negative/positive electrodes unavoidably occur during the formation processes, thereby producing organic and/or inorganic compounds between electrode and electrolyte interphases—generally known as solid electrolyte interphase (SEI) for those formed on the negative electrode and cathode electrolyte interface (CEI) for those formed on the positive electrode.^10,11 The quality of the SEI and CEI films, being largely related to the nature and properties of electrolyte, determines the kinetics of Li-ion transition through SEI film (activation energy [E_a]) and the power of blocking electron leakage from electrode materials to electrolyte (i.e., parasitic reactions).^10,12–14 In contemporary LIBs, organic carbonates (e.g., ethylene carbonate [EC], ethyl methyl carbonate [EMC], and diethyl carbonate [DEC]) and lithium hexafluorophosphate (LiPF₆) are well established as the respective main solvents and conductive Li salt due to their unique properties (e.g., high conductivity, good electrochemical stability, etc.).^3,15 The replacement of these two electrolyte components toward the improvement of the performance of LIBs could not be easily deployed owing to the overall cost per kilowatt-hour at the industrial level. In this regard, electrolyte additives used in small quantities (usually ≤5 wt% or vol%) could effectively tailor the electrochemical properties of bulk electrolytes with small impacts on their physical properties (i.e., viscosity, ionic conductivity, etc.), and thus improving the quality of SEI and CEI films and other related features required for high-performing LIBs. Hence, the design and implementation of functional electrolyte additives have been under the spotlights in recent years, as demonstrated by the research trend plot shown in Figure 1(B).

FIGURE 1. (A) Practical electrolytes in the presence of functional additives. ФA and ФC are the negative and positive electrode work functions, respectively; Eg is the window of the electrolyte for thermodynamic stability. Adapted with permission from Ref. 9. Copyright 2010 American Chemical Society. (B) Research tendency of electrolyte additives. Data were taken from Scopus® via searching the keywords of “lithium-ion batteries + electrolytes” and “lithium-ion batteries + electrolytes + additives” on April 24, 2021. CEI, cathode electrolyte interface; HUMO, highest-unoccupied molecular orbital; LUMO, lowest-unoccupied molecular orbital

To date, various kinds of electrolyte additives have been studied for LIBs to improve their performance, for instance, compounds containing unsaturated carbon–carbon bonds (e.g., vinylene carbonate [VC]), sulfur-containing compounds (e.g., PS¹⁶ and ethylene sulfite [ES]¹⁷), nitrogen-containing compound (e.g., N,N-dimethylacetamide), and silicon-containing compounds (e.g., trimethylsilyl(fluorosulfonyl)(n-nonafluorobutanesulfonyl)imide [TMS-FNFSI]¹⁸), and so on. Among these functional electrolyte additives, sulfur-containing compounds are some of the most attractive families for LIBs owing to the reasons described below. First, the replacement of carbonyl carbon with a sulfur atom endows the compounds with lower lowest-unoccupied molecular orbital (LUMO) level energies¹⁹; therefore, sulfur-containing additives are generally more inclined to electrochemical reductions compared to organic carbonates (e.g., ca. 0.8 V vs. Li/Li⁺ for EC vs. ca. 1.5 V vs. Li/Li⁺ its sulfur analog ES^2,3,17)—a desired feature of SEI-forming additives. Second, the electronegativity of sulfur is comparable to that of carbon (e.g., 2.58 [sulfur] vs. 2.55 [carbon] in Pauling scale²⁰), which renders the organic sulfur chemistry almost as rich as classic carbon chemistry—a beneficial character for customizing the properties of electrolyte additives according to the application scenarios. Third, sulfur is abundant and easily accessible,²¹ which ensures the large-scale production of sulfur-containing compounds with affordable cost—an essential parameter to be considered by battery industry.

Although the progresses of electrolyte additives in LIBs have been scrutinized in previous review articles and book chapters,^22–26 topical discussions specifically dedicated to sulfur-containing compounds as electrolyte additives are presently not available in the literature. Incentivized by our continuous efforts on the design sulfur-containing compounds for battery applications^18,27–40 and significant progresses that have been made in sulfur-containing electrolyte additives in recent years, in this review paper, we focus on the application of sulfur-containing compounds as electrolyte additives for LIBs, aiming to provide recent advances in this field and the pros and cons of popular sulfur-containing electrolyte additives. This review paper is also anticipated to be a guideline to choose specialized additives and electrolytes in commercial LIBs, which is also beneficial for the development of next-generation battery systems.

ROLES OF SULFUR-CONTAINING ELECTROLYTE ADDITIVES IN LIBs

Generally, the functional electrolyte additives used in LIBs should show distinct traits,^2,3,22,26 to be specific, sulfur-containing electrolyte additives used in LIBs usually act as 1) SEI formers, 2) CEI builders, and 3) overcharging protectors, as schematically shown in Figure 2.

FIGURE 2. The roles of sulfur-containing compounds as electrolyte additives in lithium-ion batteries (LIBs). (A) Forming robust solid electrolyte interphase (SEI) film on the negative electrode; (B) building stable cathode electrolyte interface (CEI) film on the positive electrode; and (C) preventing cell overcharging. LCO and NMC are the abbreviations of LiCoO2 and LiNi1-x-yCoxMnyO2, respectively

For the electrolyte additives used as SEI formers on the negative electrodes (i.e., graphite and silicon), they could be preferentially and rapidly reduced before the electrochemically reductive decompositions of the main solvents (i.e., EC and EMC) and conductive Li salts (i.e., LiPF₆) commonly applied in LIBs. The resulting reaction products generated via electrochemical reductions are expected to build stable SEI films (Figure 2(A)) with 1) high electrical resistance and good Li⁺ selectivity and permeability, and 2) strong tolerance toward expansion/contraction of electrode materials during cycling.^15,22,26 Particularly, SEI-forming additives with little or even without generation of gaseous compounds upon electrochemical decompositions are highly desired, since such feature is of vital importance to suppress cell swelling and delamination of electrodes, which are detrimental to the long-term cycling stability of LIBs.^15,22,26

For the electrolyte additives used as CEI builders, they could be easily oxidized on the surface of positive electrodes (e.g., LiCoO₂, LiNi_1-x-yCo_xMn_yO₂) compared to the common electrolyte compounds (solvents and Li salts) and thereby contribute to the formation of CEI film (Figure 2(B)). Moreover, the oxidative decomposition products are generally required to 1) effectively cover and/or coordinate with active sites (naked transition metal ions) of positive electrode and 2) possess strong tolerance to further oxidation (less interphase impedance change and gas evolution) upon operation.^2,3

For the electrolyte additives used as overcharging protectors, they should own desired redox activities that ensure these molecules to be oxidized on the positive electrode once the charge voltage of the battery cells goes beyond the upper cut-off charge voltage. And, in addition, the oxidized products of overcharging protectors should be soluble in the corresponding electrolytes and then diffuse back to the negative electrode for getting reversibly reduced to their pristine states (Figure 2(C)).

Based on the basic analysis above, in the following sections, the progresses of sulfur-containing compounds are presented and discussed according to their functions as electrolyte additives in LIBs. To be noticed that some sulfur-containing compounds (e.g., cyclic sulfonates) could be reduced on the negative electrode and also be oxidized on the positive electrode (e.g., for ES, ca. 1.5 V vs. Li/Li⁺ for reduction potential⁴¹ and 3.5 V vs. Li/Li⁺ for oxidation potential in carbonate electrolyte⁴²), which endows them with dual-functionalities as electrolyte additives for LIBs (see Section 3.1 for SEI formers and Section 3.2 for CEI builders).

ADVANCES IN SULFUR-CONTAINING ELECTROLYTE ADDITIVES IN LIBs

From a chemistry point of view, the sulfur-containing electrolyte additives used in LIBs could be regrouped according to their chemical structures into sulfonates (cyclic sulfonates and chain sulfonates), sulfates, sulfites, sulfones, and others (e.g., acid anhydrides, sulfide, and sulfur-containing salts), as gathered in Schemes 1 and 2. The electrochemical performances of the cells with some representative sulfur-containing electrolyte additives are collected in Table 1.

SCHEME 1. Chemical structures of the sulfur-containing compounds, sulfonates, sulfates, and sulfites, used as electrolyte additives for lithium-ion batteries (LIBs)

SCHEME 2. Chemical structures of the sulfur-containing compounds, sulfones, and others, used as electrolyte additives for lithium-ion batteries (LIBs)

TABLE 1 Impact of sulfur-containing compounds as electrolyte additives on the electrochemical performances of LIBs

Entry	Electrolyte	Cell configuration	Voltage (V)	Capacity retention change after adding additive	References
PS
1	1.0 M LiPF6-EC/EMC (1:2, by vol) + 3 wt% PS	Gr.||LNMO	3.0–4.8	+22% (0.5C, 200th cycle, 25°C)	43
2	1.0 M LiPF6-EC/EMC (1:2, by vol) + 2 wt% PS	Gr.||LCO	3.0–4.2	+10% (0.5C, 200th cycle, 25°C)	43
3	1.0 M LiPF6-EC/EMC (1:2, by vol) + 5 wt% PS	Gr.||LR-NMC	3.0–4.6	+10% (0.5C, 80th cycle, 25°C)	44
4	1.0 M LiPF6-EC/EMC (3:7, by vol) + 2 wt% PS	Gr.||NCM111	3.0–4.2	+7% (0.2C, 50th cycle, 55°C)	45
5	1.0 M LiPF6-EC/DMC + 1 wt% PS	Gr.||LFP	2.6–3.8	+12% (0.1C, 100th cycle, 25°C)	19
6	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% PS	Gr.||LCO	2.8–4.4	+26% (1.0C, 500th cycle, 25°C)	46
7	1.0 M LiPF6-EC/DEC (3:7, by wt) + 1 wt% PS	AG||NMC622	3.0–4.2	+13% (0.5C, 200th cycle, 45°C)	47
8	1.0 M LiPF6-EC/DMC/PC (1:3:1, by vol) + 2 wt% PS	Gr.||NChM532	3.0–4.5	+3% (1.0C, 120th cycle, 20°C)	48
9	1.0 M LiPF6-EC/EMC (1:2, by vol) + 5 wt% PS	Gr.||LMO	2.8–4.2	+23% (1.0C, 180th cycle, 60°C)	49
PES
10	1.0 M LiPF6-EC/DMC + 1 wt% PES	Gr.||LFP	2.6–3.8	+18% (0.1C, 100th cycle, 25°C)	19
11	1.0 M LiPF6-EC/EMC (1:2, by wt) + 5 wt% PES	Gr.||LMO	2.8–4.2	+23% (1.0C, 150th cycle, 60°C)	50
12	1.0 M LiPF6-EC/DMC/EMC (1:1:1, by wt) + 3 wt% PES	AG||LCO	3.0–4.2	+1% (1.0C, 200th cycle, 25°C)	51
13	1.0 M LiPF6-EC/DMC/EMC (1:1:1, by wt) + 3 wt% PES	AG||LCO	3.0–4.2	+13% (1.0C, 200th cycle, 70°C)	51
14	1.0 M LiPF6-EC/EMC/DEC (3:5:2, by wt) + 2 wt% PES	Gr.||LR-NMC	2.0–4.6	+10% (0.5C, 100th cycle, 25 °C)	52
BS
15	1.0 M LiPF6-EC/EMC/PC (1:3:1, by vol) + 1 wt% BS	Gr.||LCO	3.0–4.2	+1.5% (0.1C, 100th cycle, 25°C)	53
FPS
16	1.0 M LiPF6-EC/EMC (1:2, by vol) + 2 wt% FPS	Gr.||LCO	3.0–4.2	+23% (0.5C, 300th cycle, 25°C)	54
17	1.0 M LiPF6-EC/EMC (1:2, by vol) + 2 wt% FPS	Gr.||LCO	3.0–4.2	+30% (0.5C, 300th cycle, 45 °C)	54
MMDS
18	1.0 M LiPF6-EC/EMC (1:2, by wt) + 0.5 wt% MMDS	Gr.||LCO	3.0–4.5	+38% (0.5C, 100th cycle, 25°C)	55
19	1.0 M LiPF6-EC/DMC/EMC (1:1:1, by wt) + 2 wt% VC + 2 wt% MMDS	Gr.||LMO	3.0–4.2	+23% (0.5C, 450th cycle, 50°C)	56
20	1.0 M LiPF6-EC/DMC/EMC (1:1:1, by wt) + 0.5 wt% MMDS	Gr.||LMO	2.8–4.4	+4% (1.0C, 200th cycle, 25°C)	57
PSph
21	1.0 M LiPF6-EC/DMC + 1 wt% PSph	Gr.||LFP	2.6–3.8	+10% (0.1C, 100th cycle, 25°C)	58
PSOph
22	1.0 M LiPF6-EC/DMC + 1 wt% PSOph	Gr.||LFP	2.6–3.8	+35% (0.1C, 100th cycle, 25°C)	58
PMS
23	1.0 M LiPF6-EC/DEC (3:7, by wt) + 1 wt% PMS	Gr.||LCO	3.0–4.2	+2% (50th cycle, 25°C)	59
PFPMS
24	1.0 M LiPF6-EC/EMC (1:2, by wt) + 1 wt% PFPMS	Gr.||NCM523	2.8–4.2	+17% (1.0C, 400th cycle, 25°C)	60
PTM
25	1.0 M LiPF6-EC/DEC/EMC (3:2:5, by wt) + 1 wt% PTM	Gr.||NCM622	2.8–4.2	+44% (1.0C, 300th cycle, 25°C)	61
26	1.0 M LiPF6-EC/DEC/EMC (3:2:5, by wt) + 1 wt% PTM	Gr.||NCM622	2.8–4.2	33% (1.0C, 100th cycle, −20°C)	61
27	1.0 M LiPF6-EC/DEC/EMC (3:2:5, by wt) + 1 wt% PTM	Gr.||NCM622	2.8–4.2	83% (1.0C, 100th cycle, 45°C)	61
DTD
28	1.0 M LiPF6-EC/DMC + 1 wt% DTD	Gr.||LFP	2.6–3.8	+28% (0.1C, 100th cycle, 25°C)	19
29	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% DTD	Gr.||LCO	2.8–4.4	+34% (1.0C, 500th cycle, 25°C)	46
30	1.0 M LiPF6-EC/EMC (1:2, by wt) + 0.5 wt% DTD	Gr.||NCM523	2.8–4.2	+0% (0.1C, 50th cycle, −10°C)	62
PCS
31	1.0 M LiPF6-EC/DEC/EMC (1:1:1, by vol) + 1 wt% PCS + 1 wt% FEC	SiOx||LNMO	3.5–4.9	+57% (0.2C, 100th cycle, 25°C)	63
DTDph
32	1.0 M LiPF6-EC/DMC + 1 wt% DTDph	Gr.||LFP	2.6–3.8	+26% (0.1C, 100th cycle, 25°C)	58
D-DTD
33	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% D-DTD	Gr.||NCM523	2.8–4.4	+34% (1.0C, 150th cycle, 25°C)	64
34	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% D-DTD	Gr.||NCM523	2.8–4.4	+27% (0.2C, 100th cycle, −10°C)	64
35	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% D-DTD	Gr.||NCM523	2.8–4.4	+75% (1.0C, 100th cycle, 45°C)	64
ES
36	1.0 M LiPF6 -EC/DMC/DEC(1:2:2, by vol) + 0.3 wt% ES	Gr.||LCO		+7% (100th cycle, 25°C)	42
37	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% ES	Gr.||LCO	2.8–4.4	–27% (1.0C, 500th cycle, 25°C)	46
Psi
38	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% PSi	Gr.||LCO		–7% (120th cycle, 25°C)	65
DMS
39	1.0 M LiPF6-EC/EMC (1:2, by wt) + 0.5 wt% DMS	AG||NCM523	2.8–4.2	+17% (0.1C, 50th cycle, −10°C)	62
SL
40	1.0 M LiPF6-EC/DMC (1:1, by vol) + 2 vol% SL	Gr.||NCM111	3.0–4.6	+10% (0.1C, 100th cycle)	66
DMSM
41	1.0 M LiPF6-EC/DMC/EMC (1:1:1, by wt) + 0.1 wt% DMSM	Gr.||NCM111	3.0–4.6	+19% (0.2 C, 100th cycle, 25°C)	67
PTSI
42	1.0 M LiPF6-EC/EMC/DEC (1:1:1, by wt) + 0.5 wt% PTSI	Gr.||LMO	3.0–4.3	+5% (0.5C, 100th cycle, 25°C)	68
TOSMIC
43	1.0 M LiPF6-EC/DEC (3:7, by wt) + wt% TOSMIC	Gr.||NCA	3.0–4.4	+35% (0.2C, 100th cycle, 25°C)	69
PTSS
44	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% PTSS	Gr.||NCM811	3.0–4.4	+15% (1.0C, 100th cycle, 25°C)	70
NTESA
45	1.0 M LiPF6-EC/EMC (1:2, by vol) + 0.25 wt% NTESA	Gr.||NCM811		+3% (1.0C, 100th cycle, 25°C)	71
MDFA
46	1.0 M LiPF6-EC/EMC (3:7, by wt) + 1 wt% MDFA	Gr.||LCO	3.0–4.5	+21% (1.0C, 100th cycle, 25°C)	72
SPA
47	1.0 M LiPF6-EC/DMC + 1 wt% SPA	Gr.||LFP	2.6–3.8	+34% (0.1C, 100th cycle, 25°C)	19
DPDS
48	1.0 M LiPF6-EC/EMC (1:2, by wt) + 1 wt% DPDS	Gr.||LCO	3.0–4.2	+21% (1.0C, 200th cycle, 25°C)	73
49	1.0 M LiPF6-EC/EMC (1:2, by wt) + 1 wt% DPDS	Gr.||LCO	3.0–4.4	+7% (1.0C, 200th cycle, 25°C)	73
50	1.0 M LiPF6-EC/DEC (1:3, by wt) + 1 wt% DPDS	Gr.||LR-NMC	3.0–4.4	+24% (1.0C, 100th cycle, 55°C)	74
LiFSI
51	1.0 M LiPF6-EC/EMC (1:2, by wt) + 5 wt% LiFSI	Gr.||NCM532	2.8–4.4	+30% (1.0C, 100th cycle, 25°C)	75
52	1.0 M LiPF6-EC/EMC (1:2, by wt) + 5 wt% LiFSI	Gr.||NCM532	2.8–4.4	+34% (1.0C, 100th cycle, 45°C)	75
53	1.0 M LiPF6-EC/EMC (3:7, by vol) + 0.2 M LiFSI	Gr.||LCO	3.0–4.3	+20% (0.2C, 50th cycle, −20°C)	76
54	1.0 M LiPF6-EC/PC/DEC (3:4:13, by vol) + 1 wt% LiFSI	LTO||LMO	1.7–2.8	–2% (1.0C, 600th cycle, 45°C)	77
LiTFSI
55	1.0 M LiPF6-EC/PC/DEC (3:4:13, by vol) + 1 wt% LiTFSI	LTO||LMO	1.7–2.8	–3% (1.0C, 600th cycle, 45°C)	77
LiDMSI
55	1.0 M LiPF6-EC/DEC (1:1, by wt) + 2 wt% LiDMSI	Gr.||NCM111	2.8–4.5	+6% (1.0C, 100th cycle, 20°C)	78

Abbreviations: BS, 1,4-butane sultone; DTD, 1,3,2-dioxathiolane-2,2-dioxide; DMSM, di(methylsulfonyl) methane; DMS, dimethyl sulfite; DPDS, diphenyl disulfide; DTDPh, 1,3,2-benzodioxathiole 2,2-dioxide; D-DTD, dihydro-1,3,2-dioxathiolo [1,3,2]dioxathiole-2,2,5,5-tetraoxide; ES, ethylene sulfite; FPS, fluoropropane sultone; LiFSI, lithium bis(fluorosulfonyl)imide; LiTFSI, lithium bis(trifluoromethanesulfonyl)imide; LiDMSI, lithium-cyclo-difluoromethane-1,1-bis(sulfonyl)imide; MMDS, methylene methanedisulfonate; MDFA, methyl 2,2-difluoro-2-(fluorosulfonyl) acetate; NTESA, N,N,N,N-tetraethyl sulfamide; PCS, 1,3-propanediolcyclic sulfate; PS, 1,3-propane sultone; PES, prop-1-ene-1,3-sultone; PSPh, 3H-1,2-benzoxathiole 2,2-dioxide; PSOph, 1,4,2-benzodioxathiine 2,2-dioxide; PMS, propargyl methanesulfonate; PFPMS, 2,3,4,5,6-pentafluorophenyl methanesulfonate; PTM, phenyl trifluoromethane sulfonate; PSi, propylene sulfite; PTSS, phenyl trans-styryl sulfone; PTSI, p-toluenesulfonyl isocyanate; SL, sulfolane; SPA, 3-sulfopropionic anhydride; TOSMIC, p-toluenesulfonylmethyl isocyanide; Gr., graphite; AG, artificial graphite; LTO, Li₄Ti₅O₁₂; MCMB, graphitic mesocarbon microbeads; LCO, LiCoO₂; LFP, LiFePO₄; LNMO, LiNi_0.5Mn_1.5O₄; LR-NMC1, 0.5Li₂MnO₃^.0.5LiNi_0.4Co_0.2Mn_0.4O₂; LR-NMC2, Li_1.2Mn_0.55Ni_0.15Co_0.1O₂; LR-NMC3, Li_1.2Mn_0.54Ni_0.13Co_0.13O₂; LMO, LiMnO₂; NCM111, LiNi_1/3Co_1/3Mn_1/3O₂; NCM523, LiNi_0.5Co_0.2Mn_0.3O₂; NCM532, LiNi_0.5Co_0.3Mn_0.2O₂; NCM622, LiNi_0.6Co_0.2Mn_0.2O₂; NCM811, LiNi_0.8Co_0.1Mn_0.1O₂; NCA, LiNi_0.8Co_0.15Al_0.05O₂.

Sulfur-containing electrolyte additives as SEI formers

Though sulfur-containing compounds have been widely utilized as electrolyte additives to effectively passivate negative electrodes in the state-of-the-art liquid electrolytes, the research history of these additives in the battery field seems to be bewildering, Figure 3 shows the evolution of battery electrolytes containing several representative sulfur-containing electrolyte additives, including PS, PES, DTD, and ES (see Scheme 1 for chemical structures). In the early 1980s, Maxfield et al.⁷⁹ disclosed in a US patent that sultones (e.g., BS, PS) and sulfonates could sufficiently improve the stability of polymeric electrode materials such as polyacetylene (PA), polyphenylene (PPP), and poly(phenylene sulfide) (PPS) during battery cycling. The authors claimed that reducing the electrochemical potential of polymeric electrodes to be lower than 1.0 V versus Li/Li⁺ could prompt the reduction of sultones and thus achieve adequate surface coating on the electrodes. A few years after this seminal report, Simon et al.⁸⁰ evaluated several kinds of electrolyte additives including PS to suppress the self-discharge of Li metal||carbon tissue cell by minimizing the solubilization of passivation layers. Hamamoto et al.⁸⁷ from UBE Industries found that adding 0.1–4 wt% sultone could prevent the electrochemical co-intercalation of propylene carbonate (PC) into natural graphite and other highly crystalline carbonaceous negative electrode materials, thereby improving the cyclability of graphite-based cells in EC-free electrolytes. For example, the natural graphite||LiCoO₂ cells using 1.0 M LiPF₆-PC/DMC (1:1, by wt) were impossible to charge and discharge, while the addition of 0.1 wt% PS endowed the cell to be cycled for 50 cycles with 82% capacity retention. These early studies have laid a good foundation on the implementation of sulfur-containing compounds as SEI-forming additives in LIBs.

FIGURE 3. Historical evolution of the four important sulfur-containing compounds, 1, 3-propane sultone (PS), prop-1-ene-1,3-sultone (PES), 1,3,2-dioxathiolane-2,2-dioxide (DTD), and ethylene sulfite (ES) as solid electrolyte interphase (SEI)-forming additives for lithium-ion batteries (LIBs). The complete citations for the seminal patents and papers related to these electrolyte additives are accessible in the literature.17,79–86

In the 1990s, the strong interest in replacing Li metal electrode with carbonaceous materials at the negative electrode side due to safety issues stemmed from Li dendrites has largely incentivized the research on SEI-forming additives for negative electrodes.^8,26,88–91 During this blooming period of screening suitable SEI-forming additives, the other three electrolyte additives (i.e., PES, DTD, and ES) debuted in battery domain, as shown in Figure 3. In the following section, special attention is paid to these four representative sulfur-containing electrolyte additives (i.e., PS, PES, DTD and ES, see subsections of 3.1.1), and some newly emerged ones (e.g., methanedisulfonate [MMDS], lithium bis(fluorosulfonyl)imide (Li[(FSO₂)₂N], LiFSI), see subsections of 3.1.1) of particular interest for LIBs.

Representative sulfur-containing SEI formers

Instead of following the timeline shown in Figure 3, we first present individually the development of these representative sulfur-containing SEI formers (i.e., PS, PES, DTD, and ES) in the order of sultones (i.e., PS and PES), sulfate (i.e., DTD), and sulfite (i.e., ES), aiming to better reveal the chemistry of each electrolyte additives. Subsequently, a transversal comparison among these four SEI-forming additives is provided.

1,3-Propane sultone

As shown in Scheme 1, a number of cyclic sulfonates (sultones) and chain sulfonates have been investigated in the past 30 years. Amid these SEI-forming compounds, PS and PES are the two representative candidates showing features of great interest for battery applications. In typical electrolyte solutions containing 1.0 M LiPF₆ and carbonate solvents (e.g., PC, EC, and EMC), the electrochemical reduction of PS on graphite electrode normally occurs at higher potentials than those of other electrolyte components, for example, ca. 0.7 V versus Li/Li⁺ for the electrolyte of 1.0 M LiPF₆-PC/EC/EMC (1:1:3, by wt) with 1 wt% PS versus ca. 0.5 V versus Li/Li⁺ for the reference electrolyte⁸⁴ ca. 0.9 V versus Li/Li⁺ for 1.0 M LiPF₆-EC/EMC (1:2, by vol) with 2 wt% PS versus ca. 0.6 V versus Li/Li⁺ for the plain electrolyte⁴³ (Figure 4(A)). This ensures PS to be electrochemically reduced on graphite electrode prior to the decomposition or co-intercalation of solvents and salt anions, and thus the building of relatively stable SEI layers on the negative electrodes, which could prevent the degradation of electrolyte components upon continuous cycling. Introducing small amounts of PS (<5 wt%) was found to kinetically accelerate the intercalation of Li⁺ ions into graphite electrode and suppress metallic Li deposits on the surface of the negative electrode, particularly at lower temperatures (Figure 4(B)).⁹² Comparative studies on PS and VC as SEI-forming additive carried out by Lucht et al.⁴⁵ suggested that the addition of either additive effectively leads to a better capacity retention of graphite||LiNi_1/3Mn_1/3Co_1/3O₂ cells cycled at 55°C. However, the incorporation of PS resulted in the formation of lithium alkylsulfonate (RSO₂Li) without releasing gaseous compounds, differing from that of VC, which generated sizable amounts of CO₂ upon electrochemical reduction on the graphite negative electrode. This renders additional benefits on the utilization of PS as SEI-forming additives.

FIGURE 4. 1, 3-propane sultone (PS) as solid electrolyte interphase (SEI) formers for negative electrodes. (A) The first sweep in the cyclic voltagrams (enlarged view) of Li||graphite coin cells in 1.0 M LiPF6-EC/EMC (1:2, by vol) solutions containing no additive, 2 wt% SA, 2 wt% PS, and 2 wt% VC, respectively, at 25°C. Scan rate: 0.05 mV s−1. Reproduced with permission from Ref. 43. Copyright 2007, Elsevier. (B) Ex situ X-ray diffraction (XRD) patterns of natural graphite, (i) pristine natural graphite sheet, (ii) natural graphite electrode cycled in the absence of the electrolyte additive at −5°C, and (iii) natural graphite electrode cycled in the presence of the electrolyte additive at the third cycle at room temperature, followed by one discharge process at −5°C. Reproduced with permission from Ref. 92. Copyright 2009, Elsevier. (C) Thickness changes during storage of graphite||LiNi0.5Mn1.5O4 cells with 1.0 M LiPF6-EC/EMC (1/2, by vol) solutions containing no additive, 2 wt% SA, 3 wt% PS, and 2 wt% SA + 3 wt% PS, respectively. Reproduced with permission from Ref. 43. Copyright 2007, Elsevier. (D) Possible reaction paths for the reduction of PS on graphite electrodes. Reproduced with permission from Ref. 84. Copyright 2007, The Electrochemical Society

Co-utilization of PS with other electrolyte additives represents a facile approach to further enhance the cycling performance of LIBs.^{43,47,93–95} Cho et al.⁴³ reported that the co-use of PS (3 wt%) and succinic anhydride (SA, 2 wt%) remarkably suppresses irreversible gas-evolving reactions stemming from electrode–electrolyte degradations (Figure 4(C)), thus improving the capacity retention of the 5-V-class graphite||LiNi_0.5Mn_1.5O₄ pouch cells (80% capacity remaining after 100 cycles for a 50 mAh cell) at room temperature. In addition to carbonate-based electrolytes, PS has been also used as electrolyte additives for glyme- and ionic-liquid-based electrolytes for forming stable SEI films on various kinds of negative electrodes (e.g., graphite⁹⁶ and Li₄Ti₅O₁₂⁹⁷).

The chemistry behind the improved electrochemical performances of the cells using the PS-added electrolytes was discussed by Li et al.⁸⁴ The authors suggested that the heterolytic ring-opening process of PS with one electron results in the formation of the primary active species, a radical alkyl sulfonate (2, Figure 4(D)), which could be transformed into several products including alkenyl sulfonate (3, Figure 4(D)), alkyl sulfonate (4, Figure 4(D)), and sulfonate dimers (5, Figure 4(D)), as illustrated in Figure 4(D). Detailed density functional theory (DFT) calculations performed by Leggesse and Jiang⁹⁸ revealed that the termination reactions of radical alkyl sulfonates also are likely to further yield SEI-favorable species such as Li₂SO₃, which could build an effective and stable SEI film on the negative electrodes. Furthermore, with an iterative method based on DFT calculations, Hsu et al.⁹⁹ showed that the reduction of PS, in EC-based carbonate solutions, takes place faster than that of EC, and consequently the formed products could effectively suppress the continuous decompositions of EC and prevent the generation of ethylene gas (i.e., the swelling of battery cells). It has to be accentuated that the components of SEI should be chemically and electrochemically stable, thus preventing further degradations of electrolyte on the negative electrode.³ In this sense, the alkenyl sulfonate (3, Figure 4(D)) with reactive C═C bond is unlikely to be the final decomposition product of PS.

Prop-1-ene-1,3-sultone

Prop-1-ene-1,3-sultone (PES), an unsaturated cyclic sulfonate known from the late 1950s,¹⁰⁰ was initially proposed as an electrolyte additive for LIBs in 2002 by Hibara et al. from Mitsui Chemicals.¹⁰¹ The main driving force for selecting PES as an alternative SEI-forming additive to PS would be its stronger tendency toward electrochemical reduction endowed by the C═C double bond adjacent to the sulfonate (SO₃) moiety. This is effectively a good learning lesson from the implementation of the most typical unsaturated organic carbonate VC as electrolyte additive for passivating graphite negative electrode, since VC with a C═C double bond has been identified to be more prone to electrochemical reduction than its analog EC since 1997 (e.g., 0.9 V [VC] vs. 0.8 V [EC] vs. Li/Li⁺¹⁰²). In 2012, Li et al.⁸⁶ reported that the use of PES as an electrolyte additive instead of PS could more effectively improve the performance of the graphite||LiCoO₂ cells with a PC-based electrolyte (i.e., 1.0 M LiPF₆-PC/EMC [1:1, by wt]) due to the greatly suppressed co-intercalation of PC into graphite negative electrode. The experiments of cyclic voltammetry suggested that the electrochemical reduction of PES occurs at a much higher potential than that of PS (ca. 1.2 V [PES] vs. 0.7 V [PS] vs. Li/Li⁺ in 1.0 M LiPF₆-PC/EMC [1:1, by wt]), allowing PES to be preferentially reduced on graphite negative electrode prior to the co-intercalation of PC (ca. 1.0 V vs. Li/Li⁺).⁸⁶ Xia et al.¹⁰³ reported the impact of PES, PS, and VC additives on the performance of the graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells and revealed again that PES is reduced at the highest potential among these three additives, that is, ca. 1.2 V (PES) > ca. 0.9 V (PS and VC) > ca. 0.8 V (EC) versus Li/Li⁺ (being calculated according to the dQ/dV results shown in Figure 5(A)). In addition, PES can also improve the storage performance, specifically, reduce the gas release at elevated temperatures (Figure 5(B)). The SEI film formed on graphite negative electrode in 1.0 M LiPF₆-EC/EMC (1:2, by wt) with 5 wt% PES as an electrolyte additive can prevent the reduction of soluble transition metal (e.g., manganese ions) on graphite and dramatically promote the capacity retention of graphite||LiMn₂O₄ cell from 68% to 91% after 150 cycles at 60°C⁵⁰ (Table 1, entry 11). In addition, the formed SEI film in the presence of PES can effectively prevent the phase transformation of lithium-rich oxide (LRO) positive electrode via minimizing excessive electrolyte consumption, thus improving the capacity retention from 68.7% to 87.6% for the graphite||Li_1.2Mn_0.55Ni_0.15Co_0.1O₂ cells with 1.0 M LiPF₆-EC/EMC/DEC (3:5:2, by wt) containing 2 wt% PES⁵² (Figure 5(C)) (Table 1, entry 14). Further investigation revealed that the reduction mechanism of PES is quite similar to that of PS. At the first step of the reduction of PES, a radical anion coordinated with Li⁺ ion (Figure 5(D)) is formed via one electron reduction, and the chelate (1, Figure 5(D)), could be reduced and produced to intermediate (2, Figure 5(D)). Finally, intermediate (2, Figure 5(D)) could further decompose into Li₂SO₃ (3, Figure 5(D)), alkyne sulfonate (4, Figure 5(D)), alkyl sulfonate (RSO₃Li) (5, Figure 5(D)), and alkenyl sulfonate (6, Figure 5(D)). Yet, the alkyne (4, Figure 5(D)) and alkenyl sulfonates (6, Figure 5(D)) tend to be chemically and electrochemically unstable on the surface of the negative electrode due to the presence of reactive C═C and C≡C bonds, and therefore these species are unlikely to be the final products for the electrochemical decompositions of PES (cf. Section 3.3.1.1).

FIGURE 5. Prop-1-ene-1,3-sultone (PES) as solid electrolyte interphase (SEI) formers for the negative electrodes. (A) The cell terminal voltage versus capacity during the first 10 mAh and differential capacity (dQ/dV) versus voltage (V) during the formation process for the 225 mAh graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells with different concentrations of VC, PS, and PES (the control group: 1.0 M LiPF6-EC/EMC [3:7, by wt]). Reproduced with permission from Ref. 103. Copyright 2014, The Electrochemical Society. (B) Radar plots summarizing the impact of 2 wt% VC and 2 wt% PES as electrolyte additive on the performance of graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells. The axes are normalized to the worst value being equal to 100% and include coulombic inefficiency (CIE = 1 − CE). Reproduced with permission from Ref. 103. Copyright 2014, The Electrochemical Society. (C) Cyclic stability of graphite||Li1.2Mn0.55Ni0.15Co0.1O2 full cells with and without PES as an additive (standard electrolyte [STD]: 1.0 M LiPF6-EC/EMC/DEC [3:5:2, by wt]). Reproduced with permission from Ref. 52. Copyright 2020, Wiley. (D) Possible reaction paths for the reduction of PES on graphite electrodes. Reproduced with permission from Ref. 104. Copyright 2013, Elsevier

1,3,2-Dioxathiolane-2,2-dioxide

1,3,2-Dioxathiolane-2,2-dioxide (DTD), as an electrolyte additive, was introduced into Li metal cells by Mori (Mitsubishi Chem. Corp.) in 1998⁸³ and later confirmed by Hibara et al. (Mitsubishi Chem. Corp.) in 2002.⁸³ Following this seminal work, Sano and Maruyama⁸⁵ reported the use of DTD as electrolyte additive in LIBs and found that exfoliation of graphite in the PC-based electrolyte (1.0 M LiPF₆-EC/PC [1:1, by vol]) can be successfully suppressed by introducing 5 wt% DTD, owing to the high reduction potential of DTD on graphite electrode, for example, ca. 1.3 V versus Li/Li⁺ in 1.0 M LiPF₆-EC/DMC (1:1, by wt)^19,105,106 (Table 1, entry 5). However, the content of DTD used in PC-based electrolytes significantly affects the formation of SEI films on graphite negative. For example, in case of DTD content less than 3 wt% (e.g., 1.0 M LiPF₆-PC/DMC/EMC (1:1:1, by wt), the continuous reduction of PC could not be completely blocked.¹⁰⁵ Xia et al.¹⁰⁷ investigated the impact of DTD, PS, and VC as electrolyte additives on the performances of graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells (225 mAh) and observed that DTD is reduced at the highest potential among these three additives (i.e., ca. 1.3 V [DTD] > 0.9 V (VC) vs. Li/Li⁺, being calculated from dQ/dV results of the full cell, Figure 6(A)).

FIGURE 6. 1,3,2-dioxathiolane-2,2-dioxide (DTD) as an electrolyte additive for negative electrode. (A) Differential capacity (dQ/dV) versus voltage (V) during the early stages of formation or the graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells with different concentrations of VC, DTD. Reproduced with permission from Ref. 107. Copyright 2014 The Electrochemical Society. (B) Properties of graphite||LiCoO2 batteries after storage for 4 h at 85 °C. Reproduced with permission from Ref. 65. Copyright 2016, Elsevier. (C) Reduction mechanism of DTD on negative electrode. Reproduced from Ref. 108. Copyright 2017. The Electrochemical Society

The reduction chemistry of DTD on the negative electrode was investigated via DFT calculations by Hall and co-authors.¹⁰⁸ In detail, DTD tends to be reduced via a two-electron electrochemical reduction pathway and produces LiO₂SOCH₂CH₂OLi (Li₂DTD, 1, Figure 6(C)), the initial product Li₂DTD (1) could be further reduced to Li₂SO₄ and ethylene during SEI-forming processes. Li₂DTD (1) can also react with common organic solvents (i.e., EC and EMC) used in commercial LIBs, even with DTD, thus, producing organic lithium sulfite esters (ROSO₂Li, R = alkyl group, 2, Figure 6(C)). The formation of SEI films on graphite electrode in LIBs in the presence of DTD is able to enhance its cycling performance,^105,106 the initial coulombic efficiency,^65,105,109 low-temperature performance (Table 1, entry 6)⁴⁶ and suppress gas generation/explosion during storage at high temperatures.⁶⁵ For instance, Yin et al.⁶⁵ revealed that DTD improves the initial CE from ca. 90% to ca. 93% for the graphite||LiCoO₂ cell with 1.0 M LiPF₆-EC/DMC/EMC (1:1:1, by vol) and mitigates gas generation/explosion during storage at elevated temperature (i.e., 85°C, 4 h, Figure 6(B)) (Table 1, entry 38). Dahn et al.¹⁰⁶ demonstrated that the SEI film contributed by the decomposition of DTD is not stable enough for long-term cycling at 55°C (in case where DTD was used alone, the capacity of graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells with 1.0 M LiPF₆-EC/EMC [3:7, by vol] + 2 wt% DTD dropped quickly after only 150 cycles at 55°C and 0.4C); however, the co-use of 2 wt% VC and 2 wt% DTD endowed the cells with better cycling stability (i.e., 750 [VC + DTD] vs. 650 [VC] vs. ca. 400 cycles [DTD] at 80% capacity retention). In another example, Yang et al.⁴⁶ showed that, at a low temperature of 0°C, the reversible capacity of the Li||graphite half-cells containing 1.0 M LiPF₆-EC/EMC (3:7, by wt) is enhanced through the addition of 1 wt% DTD under 0.2C (e.g., 274 mAh g⁻¹ [DTD-added] vs. 187 mAh g⁻¹ [DTD-free] for the discharge capacity at cycle 50), owing to improved conductivity of electrolyte assisted by DTD. These studies are clear proofs of the beneficial impact with the inclusion of DTD as electrolyte additives for graphite-based cells.

Interestingly, in addition to graphite negative electrode, DTD endows high compatibility toward silicon oxide (SiO_x)-based negative electrode.^110,111 Zheng et al.¹¹⁰ reported that the DTD-containing electrolyte, that is, 1.0 M LiPF₆-EC/EMC/DMC/fluoroethylene carbonate (FEC) (1/1/1/0.1, by vol) + 3 wt% DTD, shows higher initial Coulombic efficiency of 83.7% from 79.4% (the control group) and high cycling performance (the Li||SiO_x cells exhibits good cycle stability after 500 cycles with 80% capacity retention). Li et al.¹¹¹ demonstrated that the addition of 1.5 wt% DTD in a well-formulated baseline electrolyte (i.e., 1.0 M LiPF₆-EC/EMC/DEC [3:5:2, by wt] + 0.5 wt% lithium difluoro(oxalato)borate [LiDFOB] + 2 wt% LiFSI + 2 wt% FEC + 1 wt% PS) results in a clear improvement in the cyclability of SiO_x-based composite||LiNi_0.8Mn_0.1Co_0.1O₂ pouch cell (e.g., ca. 560 cycles for DTD-added cell vs. 400 cycles for reference cell at the threshold value of 80% for capacity retention) due to the formation of less resistive SEI films on the negative electrode.

Ethylene sulfite

Ethylene sulfite (ES), being structurally similar to EC (see Scheme 1), is borrowed from Li metal batteries, as revealed by the early patents filled by Hoffmann.⁸¹ ES was proposed as an electrolyte solvent for secondary batteries by Ikeda et al.¹¹² (Asahi Glass Co. Ltd) in 1994. Later on, ES was suggested as an electrolyte additive for LIBs by Naruse et al. (Sony Corp.) in 1997.⁸² In view of its potential applications whatever as a solvent and/or additive in Li metal batteries and LIBs, ES was introduced into LIBs and well investigated in various literatures. In 1999, Winter et al.^17,113 reported ES as an SEI builder and found that graphite exfoliation in PC-based electrolyte is successfully suppressed, for example, in 1.0 M LiClO₄-PC/ES (95:5, by vol), ascribed to 1) the higher reduction potential of ES ahead of PC, for example, ca. 1.8 V versus Li/Li⁺ in 1.0 M LiClO₄-PC¹¹⁴ and ca. 1.5 V versus Li/Li⁺ in 1.0 M LiPF₆-PC,⁴¹ and 2) low LUMO energy of ES (−0.77 eV) compared with that of PC (0.84 eV).¹¹⁵ A step further, Xia et al.¹¹⁶ comparatively investigated electrochemical behaviors of ES and VC in graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells and revealed that ES is reduced at a higher potential, that is, ca. 1.7 V (ES) > ca. 0.9 V (VC) > ca. 0.8 V (EC) vs. Li/Li⁺, coming from the reduction potentials in the orders: 1.8 V (ES) < 2.6 V (VC) < 2.7 V (EC) from dQ/dV results (Figure 7(A)). The satisfactory SEI formed on the negative electrode in the presence of ES is highly related to electrochemical reactions therein. Leggesse and Jiang¹¹⁷ revealed that ES can be reduced via one- and/or two-electron mechanism, 1) transition state (1, Figure 7(D)) forming via one-electron electrochemical reduction and further reduction producing Li₂SO₃ (2, Figure 7(D)) and ethylene gas, and/or 2) the transition state (1, Figure 7(D)) could further react with carbonate solvent (e.g., PC) and ES, and producing the products, inorganic, Li₂SO₃ (2, Figure 7(D)), or organic (CH₂OSO₂Li)₂ (3, Figure 7(D)), and CH₃CH(OSO₂Li)CH₂OCO₂Li (4, Figure 7(D)).¹¹⁷ Moreover, the cycling performance for LIBs with ES significantly benefits from functional SEI films. Yu et al.⁴² reported that adding 0.3 wt% ES in 1.0 M LiPF₆-EC/EMC/DEC (1:2:2, by vol) endows the natural graphite (NG)||lithium cobalt oxide (LiCoO₂) cells with an increased capacity retention from 83% (ES-free cells) to 90% after 100 cycles (Table 1, entry 36). Note that Itagaki et al.¹¹⁸ reported that Li||graphite cells with ES-added electrolyte (i.e., 1.0 M LiPF₆-EC/EMC (3:7, by wt) + 1 wt% ES) show a higher value of R_SEI (ca. 85 Ω cm²) and lower coulombic efficiency (CE, ca. 65%) at the first cycle than that without ES (R_SEI: 58 Ω cm²; CE: ca. 90%). Xia et al.^116,119 observed that combining ES with VC in the graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells results in suppressed gas evolution during formation and significantly improved cycling performance. For example, the gas evolved decreased from ca. 0.9 ml for the additive-free reference cell to ca. 0.2 ml for the cell with 2 wt% VC + 1 wt% ES^116,119 (Figure 7(B)), and the capacity retention increased from 78% for the reference cell to 96% for the cell with 2 wt% VC + 1 wt% ES after 500 cycles at 1C¹¹⁶ (Figure 7(C)).

FIGURE 7. Ethylene sulfite (ES) as an electrolyte additive for negative electrode. (A) Differential capacity (dQ/dV) versus voltage (V) during the early stages of formation for the graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells with different concentrations of VC and ES (the control electrolyte: 1.0 M LiPF6-EC/EMC [3:7, by wt]). Reproduced with permission from Ref. 116. Copyright 2014, The Electrochemical Society. (B) Gas volume versus capacity during the first cycle of the formation process. Reproduced with permission from Ref. 116. Copyright 2014, The Electrochemical Society. (C) Capacity and difference between average charge and discharge voltage (delta V) (vs. cycle number) of graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells cycled at a C-rate of 1C and 40°C with VC and ES. Reproduced with permission from Ref. 116. Copyright 2014, The Electrochemical Society. (D) Possible reduction mechanism of ES. Reproduced with permission from Ref. 117. Copyright 2012, American Chemical Society

Comparison of these four electrolyte additives as SEI formers

Prior to presenting other sulfur-containing additives, the general features of these four electrolyte additives, PS, PES, DTD, and ES, discussed above, are summarized in Table 2 for a transversal comparison among these important electrolyte additives.

TABLE 2 Several parameters of the four representative sulfur-containing electrolyte additives (PS, PES, DTD, and ES) and EC, VC as a baseline

Additives/Solvents	Reduction potential versus Li/Li+/V	Advantages	Disadvantages
PS	0.9 (1.0 M LiPF6-EC/EMC [1:2, by vol])43 0.74 (1.0 M LiPF6-EC/DMC)19	High storage performance	Unstable under moisture condition, toxic and produces toxic compound120
PES	1.2 (1.0 M LiPF6-PC/EMC [1:1, by wt]),86 1.12 (1.0 M LiPF6-EC/DMC)19	Higher reduction potential and reduction kinetics than PS on negative electrode; Low interphase impedance and less gas generation during high temperature storage	Large interphase impedance upon cycling in high-voltage system
DTD	1.25 (1.0 M LiPF6-EC/DMC [1:1, by wt]),105 1.05 (1.0 M LiPF6-EC/DMC)19	Improving cycle life and storage performance	Instability during storage and large gas evolution during formation
ES	1.8 (1.0 M LiClO4-PC),114 1.541	Improve ionic conductivity; Fast reduction kinetics during SEI formation	Poor oxidation stability and storage performances, increasing interphase impedance upon operating
EC	0.8 (1.0 M LiPF6-EC/DEC [1:1])102	High compatibility against graphite negative electrode	Large gas generation during formation and storage; Solid at room temperature
VC	0.9 (1.0 M LiPF6-VC/DEC [1:1])102	Good SEI formation on negative electrode	Instability during storage

Abbreviations: EC, ethylene carbonate; ES, ethylene sulfite; DTD, 3,2-dioxathiolane-2,2-dioxide; PES, prop-1-ene-1,3-sultone (PES); PS, 1, 3-propane sultone; SEI, solid electrolyte interphase; VC, vinylene carbonate.

It is interesting to highlight that the four representative sulfur-containing electrolyte additives (i.e., PS, PES, DTD, and ES) are based on five-membered rings. As a fundamental concept of basic organic chemistry, five-membered rings own higher ring strains (i.e., the spatial orientation of atoms) than six-membered ones (i.e., 658 kJ mol⁻¹ [cyclopentane] vs. 653 kJ mol⁻¹ [cyclohexane] for the combustion heat per CH₂ unit¹²¹), which endows the former ones with lower chemical stabilities, thus more readily undergoing ring-opening reactions. This is effectively a superior benefit for these four cyclic sulfonates based on five-membered rings (i.e., PS, PES, DTD, and ES) being used as SEI-forming additives compared to those based on six-membered rings (i.e., 0.9 V PS vs. 0.7 V BS, vs. Li/Li⁺ for the reduction potential).^43,53 Additionally, five-membered rings having planar structures tend to improve the fluidity of ionic liquids (ILs), and therefore these compounds have also received great attention in building low-viscosity ILs for battery applications.^122–124

Generally, the electrochemical reduction potentials of these electrolyte additives in LiPF₆-based carbonate electrolytes decrease in the order of ES (ca. 1.5–1.8 V vs. Li/Li⁺) > PES, DTD (ca. 1.0–1.2 V vs. Li/Li⁺) > PS, VC (ca. 0.9 V vs. Li/Li⁺) > EC (ca. 0.8 V vs. Li/Li⁺) > PC (ca. 0.3 V vs. Li/Li⁺). Therefore, these four additives are quite effective in electrochemically forming sulfur-containing products, which later could contribute to stable SEI films on the negative electrodes. For example, Jankowski et al.¹⁹ comparatively investigated DTD, PS, and PES in 1.0 M LiPF₆-EC/DMC (1:1, by wt) and revealed that all these additives improve the cycling performance of graphite||LiFePO₄ cells in comparison with the reference cell without any electrolyte additives (Table 1, entry 10). The same work also demonstrated that PES prefers to reduce on graphite electrode, as indicated by the tendency of the reduction potential of these electrolyte additives: PES (1.12 V vs. Li/Li⁺) > DTD (1.05 V vs. Li/Li⁺) > PS (0.74 V vs. Li/Li⁺) (Table 2). This is ascribed to easier electrochemical reduction for the carbon alpha to unsaturated carbon–carbon double bond (C═C) of PES, in comparison with DTD and PS. However, the graphite||LiFePO₄ cell containing 1 wt% DTD displayed the highest capacity retention (i.e., the cells containing 1 wt% DTD, PES, and PS, showing 79.7%, 69.3%, and 63.7% capacity retention, respectively, after 100 cycles at 0.1 C), suggesting that SEI films formed in DTD tend to possess better quality (Figure 8(A))¹⁹ (Table 1, entry 10).

FIGURE 8. (A) Cycling performance at 0.1C rate of graphite||LiFePO4 full cells containing 1.0 M LiPF6-EC/DMC without and with 1 wt% of additive. Two separate cells were tested for each electrolyte, solid and dashed lines, and “×” indicates cell failure. Reproduced with permission from Ref. 19. Copyright 2018, Elsevier. (B) The fact of toxicity of PS. Reproduced with permission from Ref. 120. Copyright 2014, The Electrochemical Society

In addition to the reduction potential, the electrochemical stability against oxidation of an electrolyte additive is another important parameter. Yang et al.⁴⁶ systematically inspected three electrolyte additives, including ES, DTD, and PS in 1.0 M LiPF₆-EC/EMC (3:7, by wt%), used in high-voltage graphite||LiCoO₂ cells (4.4 V), and observed that capacity retentions of cells increase in the order of ES (ca. 86%) < control group without additive (91.1%) < PS (96.9%) < DTD (96.3%) after 100 cycles at 60°C (1.0°C). The cells containing ES as electrolyte additive displayed relatively poor performance compared with PS, PES, DTD, which is mainly related to lower compatibility toward graphite negative electrode, and poor tolerance to further oxidization on the positive electrode and/or high solubility of resulting products at elevated temperature, in spite of its faster reduction kinetics than that of DTD and PS⁴⁶ (Table 2). Note that, despite of its attractive performance used as an additive in LIBs compared with DTD, ES, and PES, PS is restricted to be applied into commercial battery systems, by the European chemical substance regulation (REACH: Regulation, Evaluation, Authorization and Restriction of Chemicals¹²⁵) owing to its potential carcinogenicity and toxicity to human health. Interestingly, Abe et al.¹²⁰ suggested that the safety concerns brought by PS could be lower than that of N-methyl-2-pyrrolidone (NMP), which is a typical but carcinogenic solvent used for electrode preparation, on the basis of the following considerations: 1) PS could rapidly hydrolyze to hydroxypropanesulfonic acid (HPSA, Figure 8(B)) under trace moisture conditions in jelly rolls and/or cell components during manufacturing processes, and HPSA, actually, shows much lower mutagenicity compared to PS in Ames test (i.e., a biological assay to evaluate the mutagenic potential of a given chemical, 48 000 [PS] vs. negative [HPSA]); and 2) the amount of PS utilized in a 18 650 cell is remarkably lower than that of NMP involved for electrode preparation.¹²⁰ These intriguing results effectively indicate that detailed toxicological studies on the sulfur-containing compounds of great interest for battery applications will be needed.

Other emerging sulfur-containing SEI formers

The aforementioned four important sulfur-containing compounds have been extensively investigated as suitable electrolyte additives in battery domain. Along with this mainstream research activities, other sulfur-containing variants have been also evaluated as SEI formers for LIBs, enlisting molecular (e.g., MMDS,¹²⁶ 1,3,2-benzodioxathiole 2,2-dioxide [DTDPh],⁵⁸ 3-fluoro-1,3-propane sultone [FPS],⁵⁴ phenyl methanesulfonate [PhMS],¹²⁷ propargyl methanesulfonate [PMS],^{128,129,175–177} sulfolane [SL],⁶⁶ and p-toluenesulfonylmethyl isocyanide [TOSMIC]⁶⁹), 1,3-propanediolcyclic sulfate (PCS),^63,131,132 and ionic-type (e.g., LiFSI, LiTFSI) sulfur-containing compounds.^130,174

For example, MMDS, chemical structure similar to PS, both of them including SO₃ group (see Scheme 1), shows potential application as an additive in LIBs. Xia et al.¹²⁶ comparatively studied three additives, VC, PS, and MMDS, in 1.0 M LiPF₆-EC/EMC (3:7, by wt) electrolyte applied in graphite||LiNi_1/3Mn_1/3Co_1/3O₂ pouch cells. It revealed that MMDS has the highest reduction potential among these three additives, for example, ca. 1.35 V versus Li/Li⁺ for MMDS > ca. 1.0 V versus Li/Li⁺ for PS, and VC > 0.8 V versus Li/Li⁺ for EC (coming from these electrochemical reduction peaks in the dQ/dV results of graphite||LiNi_1/3Mn_1/3Co_1/3O₂ full cells, Figure 9(A)). Further work unfolded that cycle life (particularly, high-voltage cell systems) and calendar life are significantly improved, when 0.3–2 wt% MMDS used as additives in LIBs.^56,128,133 For instance, in the case of adding 0.5 wt% MMDS into 1.0 M LiPF₆-EC/EMC (1:2, by wt), the capacity retention of the graphite||LiCoO₂ cells cycled in 3.0–4.5 V is significantly increased from 32% to 70% after 150 cycles¹²⁹ (Figure 9(B)). In addition, Cui et al. reported that PCS could be utilized as an SEI former in high voltage (5 V) SiO_x-C||LiNi_0.5Mn_1.5O₄ cell and found that the capacity retention increases from 43.3% to 100.4% after 100 cycles at 0.2C when 1.0 M LiPF₆ EC/EMC/DEC (1:1:1, by vol) + 1 wt% FEC + 1 wt% PCS is applied, which is ascribed to a functional SEI film comprised of sulfate species (e.g., Li₂SO₄), and organic sulfite species (e.g., ROSO₂Li) formed on the SiO_x-C electrode (Figure 9(C))⁶³ (Table 1, entry 31). The co-utilization of PCS and tris(trimethylsilyl)phosphite (TMSP) was found to suppress the decompositions of the co-solvent methyl acetate and graphitic mesocarbon microbead (MCMB) negative electrode, enabling the MCMB||LiNi_0.5Mn_1.5O₄ cell with good cycling performances in a wide temperature region (from −60°C to 50°C).^131,132

FIGURE 9. Methylene methanedisulfonate (MMDS) and lithium-cyclo-difluoromethane-1,1-bis(sulfonyl)imide (LiFSI) as electrolyte additives for negative electrodes. (A) The cell terminal voltage versus capacity during the first 10 mAh of the formation process and differential capacity (dQ/dV) versus voltage (V) for the nominal 220 mAh graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells with different concentrations of VC, PS, and MMDS (the control group: 1.0 M LiPF6-EC/EMC [3:7, by wt]). Reproduced with permission from Ref. 126. Copyright 2014 Elsevier. (B) Discharge capacity of cells as a function of cycle number obtained at different cut-off potentials (3.0–4.2 V and 3.0–4.5 V). Reproduced with permission from Ref. 55. Copyright 2012, Elsevier. (C) The proposed working mechanism of hybrid functional additives on SiOx electrode. Reproduced with permission from Ref. 63. Copyright 2018, American Chemical Society. (D) Capacity retention and (E) relative capacity retention of long-term cycling at 45°C with the baseline electrolyte (1.0 M LiPF6-EC/PC/DEC [15:20:65, by wt]), and baseline containing 1.0 wt% LiTFSI, 1.0 wt% LiFSI, 1.0 wt% LiBOB, 2.0 wt% LiBOB, 1.0 wt% LiDFOB, and 1.0 wt% TMSB, respectively. Reproduced with permission from Ref. 77. Copyright 2018. The Electrochemical Society

Ionic-type sulfur-containing compounds, particularly sulfonimide (SO₂N^(–)SO₂), have been long studied as conducting salts for LIBs^2,3,130,134 and Li metal batteries,^2,3 owing to their good chemical stability, structural flexibility, and anodic stability, and so on. Amid all these interesting sulfur-containing salts, LiFSI, bearing SEI-favorable FSO₂—groups, has received great attention as both main salt and electrolyte additives for LIBs.¹³⁰ The electrochemical breakdowns of FSI⁻ on the negative electrode may produce a functional SEI film consisting of LiF, Li₃SO₃, and Li₃N, which is beneficial to improve the stability of the negative electrode and facilitate Li⁺ transmission therein upon cycling.³⁶ Wang et al.¹³⁵ revealed that, when LiFSI used as co-conductive Li salt with LiPF₆ (e.g., 0.2 M LiPF₆ + 0.8 M LiFSI in EC/EMC [3:7, by wt]), the high-voltage Li||LiCoO₂ cells (4.4 V) show lower interphase impedances and less gas generation than the control group (1.0 M LiPF₆-EC/EMC [3:7, by wt]), for example, interphase resistance: 0.2 M LiPF₆ + 0.8 M LiFSI (ca. 30 Ω cm⁻²) < 1.0 M LiPF₆ (80 Ω cm⁻²) after storage 40°C for 6 h. Lucht et al.⁷⁷ comparatively investigated several ionic-type compounds, for example, LiFSI, LiTFSI, lithium bis(oxalato)borate (LiBOB), tris(trimethylsilyl)borate (TMSB), and LiDFOB, used as additives in Li₄Ti₅O₁₂||LiMn₂O₄ pouch cells (Figure 9(D,E)). When containing imide-based additives, particularly, LiFSI, the cells were endowed with higher initial coulombic efficiency and absolute discharge values at the end of cycle life (Figure 9(D)). Note that Li₄Ti₅O₁₂||LiMn₂O₄ cells containing LiFSI displayed relatively low capacity retention compared to its counterparts, for example, ca. 93.5% for TMSB < ca. 95% for 1 wt% LiFSI < ca. 95.5% for baseline < ca. 96.2% for 2 wt% LiBOB < ca. 96.6% for 1 wt% LiBOB and 1 wt% LiDFOB, owing to its relatively poor stability toward positive electrode (Figure 9(E)).

Sulfur-containing electrolyte additives as CEI builders

Interestingly, the sulfur-containing compounds can also be utilized as CEI builders in LIBs, owing to their naturally insoluble properties and robust oxidation stability of sulfates (i.e., Li₂SO₄ and ROSO₃Li), the main decomposing products formed on the positive electrode. Thereby, the usage of modified sulfur-containing compounds in LIBs can improve significantly the storage performance (capacity retention, capacity recovery, and interfacial impedance), calendar life, and suppress gas evolution upon cycling and after storage under elevated temperature.

Representative sulfur-containing CEI builders

The compounds applied as CEI builders in LIBs, as analyzed in Section 2, should show intermediate electrochemical stability compared with other electrolyte components (e.g., solvents and conductive Li salts) and their oxidative products should enable robust electrochemical stability against further oxidations toward transition metal ions on the positive electrode therein. Generally, there is a distinct difference between SEI and CEI films. As shown in Figure 10, the SEI film formed on the negative electrodes is usually as thin as 5–20 nm for graphite electrode and is slightly thicker (10–50 nm) for a silicon electrode² and could fully cover the overall surface of negative electrode under ideal condition. That is, it is that a “phase” is formed between negative electrode and electrolyte—the reason why the term “interphase” is used for SEI^2,3 In contrast, the CEI film formed on the positive electrode is not a real passivation layer and most of the surfaces between positive electrode and electrolyte are still in direct contact. For this reason, the CEI is usually called as “cathode electrolyte interface” rather than “cathode electrolyte interphase”. On the positive electrode side, the decomposition products of electrolyte additives could actively participate in the passivation of the active sites (i.e., transition metal ions not fully coordinated and exposed to electrolyte) of the positive electrode, thus decreasing the catalytic activities of transition metal ions and preventing further decompositions of electrolyte components.

FIGURE 10. Schematic illustration for the solid electrolyte interphase (SEI) and cathode electrolyte interface (CEI) films formed on the negative electrode and positive electrode, respectively

Through ab initio approaches, Kim et al.¹³⁶ demonstrated that the sulfur-containing compounds with S(═O)₂, OS(═O)₂, and OS(═O)₂O moieties can strongly stabilize Ni²⁺ in a structurally stable form owing to their moderately strong Ni²⁺-affinity, thus enhancing the structural stability of Ni-containing positive electrode materials and suppressing electrolyte decompositions. The four representative sulfur-containing electrolyte additives (i.e., PS, PES, ES, and DTD), being discussed for their usage as SEI formers in LIBs, are also of interest to be applied as CEI builders owing to their chemical structure involving OS(═O)₂ (PS, PES, and ES), or OS(═O)₂O (for DTD) (Scheme 1). Note that, on the positive electrode, the decomposition products of these sulfur-containing electrolyte additives, being mainly produced through ring-opening paths, are able to coordinate with M^2+/3+ ions to form insoluble products on CEI that could prevent the further dissolution of M^2+/3+ ions. Thus, in this section, these four representative sulfur-containing compounds (i.e., PS, PES, ES, and DTD), and some emerging CEI builders (e.g., p-toluenesulfonyl isocyanate [PTSI], dimethyl sulfone [MSM], and diphenyl disulfide [DPDS], Scheme 2) are comparably discussed.

Ethylene sulfite

In this subsection, ES is first presented as CEI builder among the four representative sulfur-containing electrolyte additives due to its lower oxidation state of sulfur atom (+IV) compared to the other three additives (i.e., PS, PES, and DTD, +VI for sulfur atom), which largely determines the electrochemical behavior of ES and the stability of the decomposition products, as illustrated below. Itagaki et al.¹³⁷ investigated the usage of ES as electrolyte additive for positive electrode by Li||LiCoO₂ half cells and observed that, with the baseline electrolyte of 1.0 M LiPF₆-EC/EMC (3:7, by vol), the ES-added electrolyte enables better oxidation stability compared to that of VC-added one (e.g., oxidation potential: VC [ca. 4.8 V vs. Li/Li⁺] < ES [ca. 5.5 V vs. Li/Li⁺]). However, the CEI film formed on LiCoO₂ surface with ES was more resistive than that of the control group without ES (e.g., ca. 30 Ω cm² [with ES] vs. ca. 10 Ω cm² [without ES] at cycle 50). The low-quality CEI film stemming from the oxidative decompositions of ES led to poor storage performance and calendar life, particularly in the high-voltage battery system. Schappacher et al.¹³⁸ systematically studied several common additives, VC, FEC, and ES, in Li||LiNi_1/3Mn_1/3Co_1/3O₂ cells and revealed that, for 1.0 M LiPF₆-EC/EMC (3:7, by wt) including 2 vol% ES, large areas of electrolyte decomposition products are formed on the aged electrode surfaces, and, moreover, the thicknesses of CEI films increase sharply under operation compared to that formed in the electrolyte containing VC or FEC (Figure 11(A)). Meanwhile, Wu et al.⁴⁶ further verified the poor storage performance of graphite||LiCoO₂ (4.4 V) pouch full cells at the elevated temperature of 85°C in the presence of ES as an electrolyte additive. The authors observed a decreased capacity retention for the cell with ES compared to the reference cell (e.g., 73.4% [1 wt% ES] vs. 86.2% [reference cell] for the capacity retentions after storage at 85°C for 12 h; Figure 11(B)). Overall, it seems that the CEI film built from the electrochemical decompositions of ES could hardly bestow the high-voltage positive electrodes (e.g., LCO, NMC, and LNMO) with sufficient cyclability upon prolonged cycles due to the poor resistance toward electrochemical oxidations of ES and ES-derived reaction products (Li₂SO₃, Li₂S₂O₃, and ROSO₂Li).⁴⁶

FIGURE 11. (A) Comparison of the cathode electrolyte interface (CEI) film thickness among different electrolyte systems (LP57 [control group] and containing 2 vol% electrolyte additives (i.e., vinylene carbonate [VC], fluoroethylene carbonate [FEC], and ethylene sulfite [ES]). Within each column, the absolute contents of the decomposition products are given, while the thickness of the unlabeled composition has a value of [less than]0.1 nm. Reproduced with permission from Ref. 138. Copyright 2016, Elsevier. (B) Capacity retention and capacity recovery ratio of graphite||LiCoO2 pouch full cells after storage at 85°C for 12 h with 1.0 M LiPF6-EC/EMC (3:7, by wt) (reference group) and containing 1 wt% DTD, 1 wt% PS, and 1 wt% ES as additives, respectively. Reproduced with permission from Ref. 46. Copyright 2018, The Electrochemical Society. (C) Full-cell performance of electrolyte with additives. Reproduced with permission from Ref. 44. Copyright 2014 The Royal Society of Chemistry. (D) Sketch of the PS protection mechanism on the Li-rich-NMC positive electrode during cycling and progressive transformation from the layered to the spinel structure. Reproduced with permission from Ref. 139. Copyright 2015, The Electrochemical Society

1,3-Propane sultone

Differing from ES, the sulfur atom in PS owns a higher oxidation state of +VI, and therefore, the decomposition products of PS would be more stable at higher potentials (>4.5 V vs. Li/Li⁺).⁴² Effectively, the role of PS on the positive electrode–electrolyte interphase was briefly discussed by Cho et al.⁴³ when analyzing the reduced swelling of graphite||LiNi_0.5Mn_1.5O₄ cells using the electrolyte containing 2 wt% PS. The authors anticipated that the stable SEI film formed on graphite negative electrode predominates over CEI in reducing the amount of gas evolution since the PS additive hardly alters the oxidation stability of the electrolyte. Yet, subsequent studies provided conclusive experimental evidences for the participation of PS in building CEI films.^{44,139–141,178} For example, Lucht et al.¹⁴⁰ suggested the R-SO₃Li species originating from the oxidations of PS contributes to forming stable CEI films on the LiNi_0.8Co_0.2O₂ positive electrode, leading to better storage performance of the MCMB||LiNi_0.8Co_0.2O₂ cell at an elevated temperature of 75°C. Kim et al.⁴⁴ also observed that PS outperforms VC in improving the cyclability of overlithiated oxides (0.5Li₂MnO₃·0.5LiNi_0.4Co_0.2Mn_0.4O₂, Table 1, entry 3, Figure 11(C)) by virtue of the coverage of alkyl sulfonate decomposition products at the surface of the positive electrode. Anouti et al.¹³⁹ reported that adding 1 wt% PS in 1.0 M LiPF₆-EC/DMC allows the formation of a stable CEI film on Li-rich-NMC particles (Figure 11(D)), which could protect the surface of active particles being exposed to electrolyte, and thus minimize the irreversible consumptions of electroactive materials by side reactions and dissolution of metal ions from Li-rich-NMC particles. A comprehensive study of the effect of various electrolyte additives (e.g., PS, FEC, VC, LiBOB, SA, etc.) in the cycling performance of Li-rich-NMC-based cells was carried out by Passerini et al.¹⁴¹ Among these studied additives, PS stood out in terms of enhancing the capacity retention and CEs of graphite||Li_0.2Mn_0.56 Ni_0.16Co_0.08O₂ cell, mainly owing to its capability in forming stable CEI film. Recent work from the same group reported the stronger effectiveness of PS in improving the cyclability of the graphite||LiNi_0.5Co_0.3Mn_0.4O₂ cell compared to FEC and VC.⁴⁸ The authors concluded that the nature of the products formed in CEI plays a more crucial role in the cycling performance of LIBs, compared to the thickness of CEI film.

Prop-1-ene-1,3-sultone

PES, similar to PS, can also be used as a CEI builder in high-voltage battery cells such as LiNi_0.5Mn_1.5O₄ (5 V)-based cells. Li et al.¹⁴³ revealed that the electrochemical oxidation of PES results in the CEI film containing Li₂SO₃ and ROSO₃Li, which could decrease the catalytic activity of transition metal ions (e.g., Mn³⁺ and Ni²⁺ in LiNi_0.5Mn_1.5O₄), and therefore suppress electrolyte decomposition on the positive electrode and enhancing the cycling performance, for example, the capacity retention of the graphite||LiNi_0.5Mn_1.5O₄ cell improved to 90% from 49% after 400 cycles at 1C when 1.0 wt% PES was added.¹⁵ Xia et al.¹⁰³ comparatively investigated the impact of VC and PES on the performance of the graphite||Li(Ni_1/3Mn_1/3Co_1/3)O₂ pouch cells and revealed that PES improves the storage performance at 60°C, for example, at the same concentration of electrolyte additive, the cells with PES showed less variation in interfacial resistance than that with VC (Figure 12(A)). However, it has to be highlighted that the unsaturated sultone PES tends to be less resistive toward oxidation than the saturated sultone PS, and generally the co-utilization of PES with other sulfur-containing additives with better oxidative stability such as sulfonates could bring beneficial impacts on cell performance. For instance, a ternary combination of 2 wt% PES, 1 wt% TMSP, and 1 wt% MMDS in 1.0 M LiPF₆-EC/EMC (3:7, by wt) endowed the graphite||LiNi_0.42Mn_0.42Co_0.16O₂ pouch cell with greatly improved capacity retention (e.g., 85% after 500 cycles vs ca. 76% after 460 cycles for the cell with PES alone).¹⁴⁶

FIGURE 12. (A) Impedance spectra change of graphite||LiNi1/3Mn1/3Co1/3O2 pouch cells with different concentrations of vinylene carbonate (VC) or prop-1-ene-1,3-sultone (PES) before and after storage at 60°C for 500 h. Reproduced with permission from Ref. 103. Copyright 2014, The Electrochemical Society. (B) Schematic illustration on the effect of PTSI as electrolyte additive for the LNMO positive electrode. Reproduced from Ref. 144. Copyright 2019, Frontiers Media S.A. (C) Electrochemical performance of Li||LiNi0.5Co0.2Mn0.3O2 cells without and with p-toluenesulfonyl isocyanate (PTSI) cycled at 1C; Reproduced with permission from Ref. 145. Copyright 2020, American Chemical Society. (D) The reaction mechanism for the oxidative polymerization of diphenyl disulfide (DPDS). Reproduced with permission from Ref. 74. Copyright 2017, Elsevier

1,3,2-Dioxathiolane-2,2-dioxide

DTD with better oxidative stability because of its highest oxidation state can be also used as electrolyte additive to improve the compatibility of electrolyte toward positive electrode materials (e.g., LiCoO₂, LiNi_0.5Mn_1.5O₄, LiNi_1-x-yCo_xMn_yO₂).^147–149 For instance, Dahn et al.¹⁴⁷ reported that the addition of DTD to graphite||LiNi_0.4Mn_0.4Co_0.2O₂ pouch cell drastically mitigates the increase of cell impedance and improves its capacity retention. The co-use of DTD and FEC allowed the graphite||LiNi_0.5Mn_0.3Co_0.2O₂ cells (4.3 V) to achieve good charging C-rates (0.3C) and long-term cycling stability (92% capacity retention after 1400 cycles at 40°C).¹⁴⁸

Comparison of these four electrolyte additives as CEI builders

In summary, the incorporation of the aforementioned four representative sulfur-containing electrolyte additives (i.e., DTD, PES, PS, and ES) favors the formation of sulfate species (i.e., Li₂SO₄ and ROSO₂Li) on the surface of positive electrode.^52,108,117 The as-formed sulfate species possess several advantages over other CEI components (i.e., RCOOLi, ROCOOLi, Li₂CO₃) originating from the decompositions of conventional electrolyte additives (e.g., FEC, VC), including (1) lower solubility in organic carbonate solvents, even under elevated temperatures, which could effectively suppress the dissolution of transition metal ions¹⁴³; (2) stronger coordination power toward transition metal ions, which could decrease the catalytic activity of active sites and prevent continuous oxidative decompositions of electrolyte components^2,3,68; and (3) possibly higher ionic conductivity, particularly when deposited as ultrathin layers around a few nanometers, which could promote the transport of Li⁺ ions across the interface between cathode and electrolyte.¹⁵⁰ Overall, PS, PES, and DTD with a higher oxidation state of the sulfur atom would be more suitable as CEI builders for high-voltage positive electrodes than ES. The anodic stability of the CEI components and the ability to stabilize transition metals exposed to electrolytes would be important indicators to assess new sulfur-containing CEI builders.

Other emerging sulfur-containing CEI builders

Along with the development of electrolytes containing above four representative sulfur-containing CEI builders, there has been extensive effort on searching other potential sulfur-containing molecules, including PTSI,^{68,144,145,151–153} MSM,¹⁵⁴ DPDS,⁷⁴ dihydro-1,3,2-dioxathiolo [1,3,2]dioxathiole-2,2,5,5-tetraoxide (D-DTD),⁶⁴ lithium-cyclo-difluoromethane-1,1-bis(sulfonyl)imide (LiDMSI),⁷⁸ methyl benzenesulfonate (MBS),⁷⁸ methyl 2,2-difluoro-2-(fluorosulfonyl) acetate (MDFA),⁷² N,N,N,N-tetraethylsulfamide (NTESA),⁷¹ 1,2,6-oxadithiane 2,2,6,6-tetraoxide (ODTO),¹⁵⁵ 2,3,4,5,6-pentafluorophenyl methanesulfonate (PFPMS),⁶⁰ phenyl trifluoromethane sulfonate (PTM),⁶¹ phenyl trans-styryl sulfone (PTSS),⁷⁰ and so on. For example, it has been confirmed that PTSI is able to form a functional stable CEI film on the positive electrode, consisting of Li₂SO₃, Li₂S, and ROSO₂Li, particularly in high-voltage systems (>4.2 V). The CEI formed on the positive electrode surface, SO₂ involved therein, delocalizing the nitrogen nucleus to block the reactivity of PF₅ toward transition metal ions in positive electrode (Mn³⁺, Ni²⁺, and Co²⁺), produced from LiPF₆ reacting with traces of protic material in cells. As result, transition metal ions dissolution is effectively suppressed, and cycle performance together with calendar life are also improved (Figure 12(B)).^{68,144,145,151–153} For instance, Li et al.¹⁵² revealed that Li||LiMn₂O₄ cells (4.35 V) at 55°C, containing 0.5 wt% PTSI in 1.0 M LiPF₆-EC/DEC/EMC (1:1:1, wt%) deliver 81% capacity retention after 100 cycles at 55°C, better than that without PTSI (with capacity retention 76% after 100 cycles). Wang et al.¹⁴⁵ reported that, when 0.5 wt% PTSI included in 2 M LiPF₆-EC/DEC (1:1, by wt), the high-voltage graphite||LiNi_0.5Co_0.2Mn_0.3O₂ cells (4.8 V) stably operates for 640 cycles with capacity retention increasing from 29% to 66% at 1 C (Figure 12(C)), and Mn³⁺ dissolution into electrolyte is effectively suppressed therein.

In another example, Zhao et al.¹⁵⁴ reported the benefit of using MSM (see chemical structure in Scheme 2) as a CEI former in Li||LiCoO₂ half-cells (4.6 V). The cell with 1.0 M LiPF₆-DMC/FEC/1,1,2,2-tetrafluoroethyl 2,2,3,3-tetrafluoropropyl ether (HFE) (30:30:35, by wt) + 10 wt% MSM showed 75% capacity retention after 300 cycles at 1 C, which was much better than the control group (ca. 21.0% capacity retention after 300 cycles for the cell with 1.0 M LiPF₆-EC/DMC [1:1, by mole]). Zuo et al.⁷⁴ reported diphenyl disulfide (DPDS) as an electrolyte additive for passivating positive electrode in graphite||Li_1.2Mn_0.54Ni_0.13Co_0.13O₂ pouch cells (4.4 V) and demonstrated that DPDS is preferentially oxidized (oxidation potential: 4.2 V and 4.6 V vs. Li/Li⁺ for 1 wt% DPDS in 1.0 M LiPF₆-EC/DEC (1:1, by wt), and 4.7 V vs. Li/Li⁺ for the control group without DPDS) and polymerized on Li_1.2Mn_0.54Ni_0.13Co_0.13O₂ positive electrode producing polymerization products with high electron conductivity (Figure 12(D)). The resulting CEI film effectively covered active sites and enhanced the conductivity of the positive electrode material, leading to good cycling performance and low interphase impedance, for example, capacity retention of the cells containing 1.0 wt% DPDS in 1.0 M LiPF₆-EC/DEC (1:1, by wt) increased from 44% to 68% after 100 cycles at 55°C.

Sulfur-containing electrolyte additives as overcharging protectors

Overcharging process of LIBs, being referred to the charging voltage applied beyond a specified value according to the chemistry of the battery cell, is an extremely dangerous behavior upon cell cycling.^2,156 It may induce continuous parasitic reactions of battery compounds on the positive electrode and dendrite growth on the negative electrode and thus leading to thermal runaway, leakage, fire, and even explosion.^2,3 Hence, eliminating the side effect of overcharging process is a severe challenge and has been cited as a compulsive requirement in UN 38.3 (Recommendations on the Transport of Dangerous Goods: Manual of Tests and Criteria, subsection 38.3¹⁵⁷) for consumer electronic products. The usage of overcharging protectors, also called redox-shuttle additives, is one of most effective methods to overcome side effects of overcharging on LIBs, particularly for personal safety of customers. Normally, the chemical compounds (abbreviated as S), potentially applied as overcharging protector in LIBs, should possess an intrinsic oxidation reaction at the charging voltage slightly exceeding the specialized cut-off voltage. And the resulting oxidation products (S^∙+) could be subsequently dissolved in electrolyte and get reduced to pristine state (S) on the negative electrode (Figure 13(A) and Figure 2(C)).

FIGURE 13. (A) Schematic illustration on the working mechanism of overcharging protectors (abbreviated as S). (B) Constant charging curves (voltage vs. time) for the cells with the electrolyte of 1.2 M LiPF6-EC/EMC containing various kinds of overcharging protectors (i.e., 0.08 M DBB, EPT, and BCF3EPT). The cells were programmed to charge at 0.1C until reaching a cell voltage of 5.0 V. Reproduced with permission from Ref. 158. Copyright 2014, The Electrochemical Society

Generally, the overcharging protectors own several specialized traits, including 1) suitable redox potential related to the specialized battery cell system, 2) enough electrochemical stability within the operational voltage of the corresponding LIBs, and 3) adequate solubility in the electrolyte solution to enable sufficient redox-reaction cycles (Figure 13(A)) during overcharging processes.

So far, overcharging protectors are mainly selected from 1) redox-active inorganic compounds (e.g., I⁻/I₂,^159,160 Br⁻/Br₂¹⁶¹), 2) coordination compounds (e.g., ferrocenes), and 3) organic compounds containing heteroatoms (nitrogen, oxygen, and sulfur¹⁶²). The inherent oxidation/reduction potentials of different overcharging protectors determine their application in specialized battery cell systems. For instance, the redox potential of I⁻/I₂ lies close to 3.1 V versus Li/Li⁺, implying that such additive would be only of potential interest for 3-V-class instead of 4-V-class LIBs.^159,160

Sulfur-containing compounds used as overcharging protectors could be mainly classified into two types, that is, 1) thianthrene and 2) phenothiazine (Scheme 2). Via changing the chemical structures of the substituent groups, the oxidation/reduction potentials of the overcharging protectors can be adjusted accordingly. Lee et al.¹⁶³ synthesized series of thianthrene-type compounds (e.g., 2, 7-diacetylthianthrene [DATT], 2, 7-dibromothianthrene [DBTT], 2, 7-diisobutanoylthianthrene [DIBTT], and 2-acetylthianthrene [ATT]; Scheme 2) and investigated their electrochemical behavior in nonaqueous electrolytes. It was found that these thianthrene-type overcharging protectors display redox potentials, being largely related to the nature of substituent groups, in the range of 4.0–4.4 V versus Li/Li⁺. For instance, thianthrene showed a redox potential of ca. 4.1 V versus Li/Li⁺; however, DBTT with an electron-withdrawing group (i.e., bromine) exhibited a higher redox potential of ca. 4.4 V versus Li/Li⁺, indicating the possible use of the latter one in typical LIBs with an upper cutoff voltage of 4.2 V.¹⁵⁴

Besides, Dahn et al.¹⁶⁴ reported series of phenothiazine-type compounds (e.g., 10-ethyl-phenothiazine [EPT], 10-ethyl-3-chloro-phenothiazine [CEPT], 10-isopropyl-phenothiazine [IPT], and 10-acetyl-phenothiazine [APT]) as overcharging protectors for Li₄Ti₅O₁₂||LiFePO₄ cells. The authors observed that incorporating 0.1 M APT (redox potential: 3.9 V vs. Li/Li⁺) in 0.7 M LiBOB-PC/DMC/EC/DEC (1:2:1:2, by wt) allows the cell to be cycled under overcharging conditions with longer life. In addition, Odom et al.¹⁵⁸ comparably studied three overcharging protectors (i.e., 1,4-di-tert-butyl-2,5-dimethoxybenzene [DBB], EPT, and 3,7-bis(trifluoromethyl)-N-ethylphenothiazine [BCF3EPT]) in 1.2 M LiPF₆-EC/EMC (3:7, by wt) and applied them into graphite||LiFePO₄ cells. It was shown that 0.08 M BCF3EFT endows the cell with the best resistance toward overcharging process, for example, 3448 h (BCF3EPT) > 1580 h (EPT) > 606 h (DBD) before a sharp rise of cell voltage to the threshold value of 5 V (Figure 13(B)).

In short, sulfur-containing compounds are of particular interest for being used as overcharging protectors for LIBs, as their redox potentials are easily tailored via introducing substituent groups into sulfur atoms.

CONCLUSIONS

Electrolyte additives have been known as “small dose, big impact”. The aforementioned advances in sulfur-containing additives clearly show that the utilization of these additives alone or together with other electrolyte additives could greatly tailor the properties of bulk electrolyte and/or electrode–electrolyte interphases and thereby improving the cyclability and energy density of LIBs. To accelerate the development of sulfur-containing electrolyte additives for battery applications, the following aspects could be tackled in upcoming research work.

Computational-assisted approaches in screening new additives

Presently, DFT-based methods have been well employed to predict the resistivity of electrolyte additives toward electrochemical reductions/oxidations via HOMO/LUMO energy levels calculations^46,84,92,115 and to search for possible reaction paths for the decomposition processes via energy calculations of possible intermediates.¹⁹ Recent studies have shown that the surrounding environment of electrolyte additives (electrolyte salt and solvent) greatly affects the predicted properties of electrolyte additives,^99,165,166 and therefore special attention has to be taken when correlating the computational results with experimental ones. Additionally, the deployment of multiscale calculations (from nanoscopic to macroscopic levels) and the combination of emerging data-driven science (i.e., machine-learning techniques) with quantum chemistry calculations are very likely to boost the design and screening of suitable electrolyte additives.^167–170

Understanding the role of electrolyte additives with suitable cell configuration

Although half-cell configurations with Li metal as negative electrode have been extensively used for investigating the properties of electrolyte additives, it has to be accentuated that the intrinsic dilemmas including the growth of dendritic Li and extremely high reactivity toward electrolyte components may bring uncertainly on the evaluation processes related to the electrolyte additive customized for LIBs rather than Li-metal-based batteries.⁴⁸

Evaluating the cross-talk between negative and positive electrodes

The diffusion of soluble SEI/CEI species to the surface of positive/negative electrode originating from the other side has been reported in recent work.^171–173 These interesting results suggest that cross-diagnosis of the interphases between negative and positive electrode and electrolyte would benefit the understanding on the chemistry of electrolyte additives in LIBs.

ACKNOWLEDGMENTS

The authors would like to thank the support from the National Natural Science Foundation of China (No. 51172083) and the Fundamental Research Funds for the Central Universities (2020kfyXJJS095).

CONFLICT OF INTEREST

The authors declare no conflict of interest.