Since the theoretical specific capacity of sulfur (1673 mAh/g) is much inferior to lithium (3860 mAh/g), the initial energy density is determined by the mass of sulfur in the whole system. Moreover, the instability at the surface and the bulk of the lithium anode during cycling seriously affect the capacity retention and cycle life [38]. According to recent progress, Li-S batteries are suffering from the following challenges and difficulties: (1) the Shuttle effect of Li ions [39]; (2) the uneven deposition of intermediate-term for SEI [40]; and (3) the serious dendrites and self-discharge of the lithium metal anode [4]. Among these factors, the serious dendrites of the lithium metal anode have the greatest impact on safety and stability [41]. The formation of lithium dendrites in lithium-ion batteries is attributed to the repeated deposition and stripping of lithium metal on the surface of electrodes during the charging and discharging process. Pore formation and surface defects can occur as a result of this process, leading to a high local potential that promotes the growth of dendrites [42]. Other factors such as temperature, current density, composition and concentration of electrolytes, electrode materials and structures, and the charging/discharging states of the battery can also contribute to dendrite formation. Lithium dendrites lead to serious safety issues, as they can cause short circuits, puncture the separator, and initiate uncontrolled reactions [43]. In order to overcome these difficulties and obtain stable lithium metal anodes, numerous strategies such as regulating the electrolyte, fabricating an artificial SEI layer, modifying the 3D current collector hosts and/or the lithophilic site, and exploring alternative anode materials have been employed [44]. However, research on Li-S batteries operating at low temperatures is still scarce.

Graphite, with its flat intercalation potential, good capacity, and long cycle life, can effectively inhibit volume expansion and dendrite formation in lithium metal anodes. However, the related intrinsic drawbacks, such as poor Li diffusion kinetics in the interlayers, low intercalation potential, and relatively large interfacial resistance [45,46], usually result in the Li plating on graphite and restricting extensive applications in low-temperature conditions [45,46,47]. Due to the catalytic effect of the metal during the Li ions desolvation process, when mixed with 1 wt% of metal nanoparticles (Cu, Al, Sn), the light graphite oxide was able to deliver 30% of the theoretical capacity at −30 °C and 0.2 °C [48,49]. In comparison with the original graphite oxide electrode, the graphite oxide with coated and dispersed Sn provided a capacity of 152 and 94 m Ah g⁻¹, respectively, at −30 °C. It was demonstrated that the low-temperature performance of graphite and LTO(Li₄Ti₅O₁₂) anodes can be effectively improved by using copper nanoparticles on Super-P (Cu/Super-P) as a conductive additive [50,51]. Coating graphite with Al₂O₃ may also prevent Li deposition, improving its cryogenic performance [52].

The mild oxidation of graphite is expected to solve the problem of reducing the lithium overpotential of the graphite anode. Cao [53] and Wu et al. [54] innovatively utilized the heat treatment and concentrated nitric acid solution to lightly oxidize graphite, which exhibits better cycling performance at low temperatures. The reduced particle size and number of terminal unsaturated carbon atoms, forming nanoscale voids and channels as well as the formation of chemically bonded SEI, was attributed to this mild oxidation treatment [47].

In addition, particle size reduction and structure modification can also improve the low-temperature performance of graphite [55,56]. The combination of thin graphite sheets with through-holes (porous graphite nanosheets, PGNs) and carbon nanotubes (CNTs) significantly shortens the diffusion path. The dominant mesopores and micropores in PGN-CNT anodes facilitate Li ion transport, resulting in a superior rate and performance at low temperatures (Figure 2a,b). After 500 cycles, the capacity retention was maintained at 90% for 0.75 m LiTFSI 1,3-dioxane (DIOX) electrolyte, and the reversible capacity was above 300 mAhg⁻¹ at 0.1 °C and −20 °C. Ionic liquids were used to perform microwave exfoliation on expanded graphite and synthesized multilayer crystalline graphene (GRAL) (Figure 2c) [57]. The high surface area allows efficient electrochemical reactions, as evidenced by the 3–4 times greater capacity of GRAL compared to commercial graphite at −30 °C (Figure 2d). Moreover, oxided mesocarbon microbeads and expanded MCMB were prepared as well. The expanded MCMB with increased interlayer distance delivered 130 and 100 m Ah g⁻¹ at −10 and −40 °C, respectively [58].

Recently, energy-dense Li-S batteries were achieved with the assistance of a gel polymer electrolyte, which was designed by redox chemistry between a Li₂S cathode and Si anode. The rationally designed CoN@MCNF/Li₂S cathode shows good temperature adaptability at around −20 °C [59].

Jiao et al. [60] systematically studied the performance of lithium metal anodes over a range of temperatures. A variety of electrolyte systems have been studied, such as a cosolvent of 1,3-dioxolane (DOL)/1,2-dimethoxyethane (DME), and a cosolvent of ethylene carbonate (EC) and dimethyl carbonate (DMC). Both drupe-like and small spherical deposits of lithium were detected at high temperatures of 60 °C. It was shown that between five degrees and minus fifteen degrees, there will be branching of lithium metal anodes. However, at higher temperatures, the C–C and C–O groups of organics in the SEI are enriched, resulting in a low modulus, flexible SEI structure. In thermodynamic and kinetically favored depositional conditions, low-current density, high Coulombic efficiency as well as a nondendritic structure has been achieved in these electrolytes.

Compared with low temperatures, conventional carbonate LiFP₆ electrolytes are preferred over poorer SEI layers, with lower Coulombic efficiency and shorter cycle life at high temperature [61]. At 0 °C, a homogenous SEI layer was formed that contained uniformly distributed F-containing species, displaying an extended cycle lifetime and increased Coulombic efficiency (Figure 3). The XRF mappings and micro-XANES images illustrate that F and O are inhomogeneously distributed over most of the surface and from the surface to the bulk, which is important to understand. In particular, when the cycle temperature reaches below 60 °C, the internal composition of the SEI is mainly composed of inorganic Li₂CO₃, and the surface composition is mainly composed of organic oxygen and LiF. These species are produced by the reaction between the electrolyte and the deposited lithium, thus causing rapid failure when building a thick SEI.

Robert [62] and co-workers demonstrated that using the positive temperature gradient method (the positive electrode is set to 40 °C and the negative electrode is set to 0 °C) could produce uniform lithium quiescence and reduce deleterious side reactions. In typical LiTFSI-LiNO₃-based ether electrolytes, after one hundred cycles, the symmetrical cell with positive thermal gradient exhibited the lowest voltage hysteresis and the longest cycle life. Whereas, under negative temperature gradient conditions (negative pole set to 0 °C and positive pole set to 40 °C), a rapid degradation of battery performance was monitored after only 15 cycles. It demonstrated that the very first lithium seeding state enabled control of quality and cycle life of Li metal anodes. In particular, negative temperature gradients can lead to cold seeding and thermal stripping processes, which generally produce inhomogeneous coatings with high aspect ratios, whereas positive temperature gradients can produce uniform initial lithium quiescence with good long-term cycling performance. These external temperature gradient methods are conducive to stabilizing Li metal anodes.

To achieve an ultra-low temperature lithium metal anode, Xia [63] and co-workers developed a cosolvent electrolyte system by adding electrochemically “inert” dichloromethane (DCM) as diluent to a concentrated ethyl acetate (EA)-based electrolyte. By adding DCM diluent, the solvated structure was changed. In EA solution, the mobile DCM molecules wrap the salt clusters of LiTFSI. Due to this unique structure, the electrolyte exhibits high ionic conductivity, low viscosity, and low-temperature performance. Although the stripping/plating performance of the Li metal anode was not shown, the Li//PI battery exhibited good cycling performance at −70 °C and good performance at low temperatures. However, based on the investigation of the Golmohammad group, the newly fabricated solid-state electrolytes show good stripping/plating performance toward the metallic Li anode under such conditions [64,65,66].

The problem of the high lithium ion desolubilization energy barrier at low temperatures can be effectively mitigated by artificially adjusting the SEI film. The slow process at the lithium anode occurs when the solvated Li attempts to enter the anode structure, where it must peel off its primary solvated sheath [67]. This dynamically challenging process, often referred to as the “charge transfer” component due to its characteristic 50–70 kJ mol⁻¹ activation energy barrier, is the fundamental reason why LIBs cannot be charged at low temperatures. However, the activation energy of the Li ions’ interfacial transfer is different in the presence of different surface films (SEI), and the dynamics of cathode–interface Li ion transfer are also affected by the composition of SEI film [68,69]. Therefore, the artificially modified SEI film effectively affects the “charge transfer” process, providing a beneficial strategy to solve the high Li ion desolubility barrier problem (Figure 4).

The 3D current collector hosts and lithophile site modification are effective to achieve high-performance lithium anodes and fast conversion kinetics [45]; their feasibility and reliability in extreme hot and cold conditions, however, have been studied relatively infrequently. Peng and co-workers [72] used chemical vapor deposition to produce graphite-coated Ni-Fe foam (graphite @Ni-Fe). It exhibited good cycle performance, and the capacity retention rate was 98.10% after 100 cycles at −50 °C and 0.4 A cm⁻², with a Coulomb efficiency over 98.3%. At present, zinc batteries are performing at low temperature, but lithium batteries are not.