Cell Design for Improving Low-Temperature Performance of Lithium-Ion Batteries for Electric Vehicles

: With the rapid development of new-energy vehicles worldwide, lithium-ion batteries (LIBs) are becoming increasingly popular because of their high energy density, long cycle life, and low self-discharge rate. They are widely used in different kinds of new-energy vehicles, such as hybrid electric vehicles and battery electric vehicles. However, low-temperature ( − 20– − 80 ◦ C) environments hinder the use of LIBs by severely deteriorating their normal performance. From the perspective of material design, this review summarized and analyzed common methods of improving LIBs’ performance via structure optimization and material optimization, and the future development of methods in this regard is discussed. This review is expected to provide cell design ideas for enhancing the low-temperature performance of LIBs.


Introduction
The world is facing an unprecedented energy crisis and environmental pollution [1,2]. The high demand for petroleum fuels in the automotive industry and the excessive emission of pollutants are among the important reasons for the global energy crisis and environmental pollution. Therefore, countries around the world are actively taking measures to develop new-energy vehicles to alleviate energy shortages and environmental pollution [3].
As a kind of new-energy vehicles, electric vehicles have an energy conversion rate of 75% or higher, which is considerably larger than that of conventional vehicles, and are characterized by energy savings and low pollution [4][5][6][7]; therefore, they are widely considered a promising approach to achieving a clean, efficient transportation system [8]. Electric vehicles have developed rapidly in recent years and have gradually become a global hot spot. Because lithium-ion batteries (LIBs) have a high specific energy, long life, excellent safety, fast-charging capability, low self-discharge, and eco-friendliness, a vehicle equipped with LIBs has a relatively long electric endurance mileage and can meet the power requirements of electric vehicles [9][10][11]. Furthermore, because LIBs can be combined with grids to store electricity from sustainable energy sources, such as photovoltaics and wind energy, LIBs have become a key part of the development of electric vehicles [10]. LIBs are also widely used to power electric vehicles because of their performance, safety, and cost advantages [12,13].
However, LIBs' performance is greatly affected by the environmental conditions, especially at low temperatures, ranging from −20 • C to −80 • C. The deterioration of LIBs' performance at low temperatures mainly manifests in the following aspects. (1) The discharge capacity reduces. Li et al. [14] found that, at −40 • C, a 1 Ah LiNi 1/3 Co 1/3 Mn 1/3 O 2 performance at low temperatures mainly manifests in the following aspects. (1) The discharge capacity reduces. Li et al. [14] found that, at −40 °C, a 1 Ah LiNi1/3Co1/3Mn1/3O2 (NCM111) pouch cell with a conventional electrolyte retained only approximately 20% of its initial discharge capacity. (2) The rate performance deteriorates. Jaguemont et al. [15] reported that, when the temperature dropped to −25 °C, the discharge rate of an LIB had to be set to 0.2 C or even less to maintain its room-temperature discharge capacity. In addition to the reduced energy and power output, there are also thermal safety issues due to Li plating. The main parameters to control or improve the performance of LIBs at low temperatures include ionic conductivity, impedance, the ion diffusion rate, and so on [16,17,18,19]. Zhang et al. [20] summarized the factors that deteriorate the performance of LIBs at low temperatures, which mainly include: (1) the decreased ionic conductivity and wettability of liquid electrolytes; (2) the increased intrinsic grain-boundary resistance and slow Li + diffusion rate in electrodes; (3) the difficult Li + dissolution and sluggish Li + transport through the solid-electrolyte interphase (SEI) along with high charge-transfer resistance; (4) the occurrence of Li plating. These factors are shown in Figure 1 from Zhang et al. [20]. The first factor leads to high viscosity or solidification of the electrolyte, which hinders the ionic transport; the second factor inhibits the (de)lithiation reactions; the third one slows down the battery kinetics; the fourth one not only consumes lithium-ions, resulting in a reduction of reversible lithium-ions, but also easily causes safety problems [21,22]. The key consideration in alleviating the capacity deterioration of LIB is the sluggish kinetics of Li + [23].  [20].
In order to improve the low-temperature performance of batteries, from the perspective of the system, researchers often focus on optimizing the battery's thermal management system to improve the temperature of the battery's operating environment [8]. From the perspective of materials and cells, researchers often focus on improving the anode, cathode, and electrolyte materials, optimizing the porosity, compaction density, and thickness of the electrode, so as to essentially improve the ionic conductivity, ionic conduction rate, and diffusion rate of the battery and shorten the ionic conduction distance [20,23]. However, there are few articles summarizing the current research status of enhancing the low-temperature performance of LIBs from the perspective of cell design, including material optimization (anode, cathode, and electrolyte) and structure optimization (tortuosity and porosity). They are shown in Figure 2. They can both increase the battery discharge capacity and improve the rate performance. The optimization of Figure 1. Factors deteriorating LIB performance at low temperatures [20].
In order to improve the low-temperature performance of batteries, from the perspective of the system, researchers often focus on optimizing the battery's thermal management system to improve the temperature of the battery's operating environment [8]. From the perspective of materials and cells, researchers often focus on improving the anode, cathode, and electrolyte materials, optimizing the porosity, compaction density, and thickness of the electrode, so as to essentially improve the ionic conductivity, ionic conduction rate, and diffusion rate of the battery and shorten the ionic conduction distance [20,23]. However, there are few articles summarizing the current research status of enhancing the low-temperature performance of LIBs from the perspective of cell design, including material optimization (anode, cathode, and electrolyte) and structure optimization (tortuosity and porosity). They are shown in Figure 2. They can both increase the battery discharge capacity and improve the rate performance. The optimization of anode and cathode materials can effectively reduce the charge-transfer resistance at low temperatures, shorten the diffusion distance of lithium-ions, accelerate the diffusion rate of lithium-ions and, then, enhance the diffusion kinetics of Li + , improve the discharge capacity of the battery, and improve the rate performance [10,17]. The design and development of the electrolyte can reduce the freezing point of the solvent, improve the ionic conductivity, and then, increase the capacity of the battery at low temperatures, which result in a considerable improvement in the discharge capacity of the LIBs at low temperatures [14,16]. Low electrode tortuosity and high electrode porosity improve the lithium-ion transport in an electrode; it can speed up the ion transport rate and shorten the lithium-ion transport path, effectively improving the lithium-ion transport efficiency [23].
Focusing on LIBs for electric vehicles, this paper summarizes and analyzed published methods of improving the low-temperature performance of LIBs from the viewpoint of cell design. The possible effect of changing the battery structure on the low-temperature performance of LIBs is discussed. Next, carbon materials, lithium titanate, layered oxides, spinel-structured oxides, and other anode and cathode materials are explained. The remarkable effects of various electrolytes (including four common materials such as aqueous electrolyte and organic electrolyte) on enhancing the low-temperature performance of LIBs are then analyzed. The organic electrolyte is discussed from three aspects: solvents, lithium salts, and additives. Finally, the development opportunities for the performance-improving methods of LIBs under low-temperature conditions are given, followed by some opinions and suggestions. This review is expected to inspire further studies on enhancing the low-temperature performance of LIBs from the perspective of cell design.  Focusing on LIBs for electric vehicles, this paper summarizes and analyzed published methods of improving the low-temperature performance of LIBs from the viewpoint of cell design. The possible effect of changing the battery structure on the low-temperature performance of LIBs is discussed. Next, carbon materials, lithium titanate, layered oxides, spinel-structured oxides, and other anode and cathode materials are explained. The remarkable effects of various electrolytes (including four common materials such as aqueous electrolyte and organic electrolyte) on enhancing the low-temperature performance of LIBs are then analyzed. The organic electrolyte is discussed from three aspects: solvents, lithium salts, and additives. Finally, the development opportunities for the performance-improving methods of LIBs under low-temperature conditions are given, followed by some opinions and suggestions. This review is expected to inspire further studies on enhancing the low-temperature performance of LIBs from the perspective of cell design.

Cell Fabrication/Structure Optimization
Battery components and fabrication methods affect the low-temperature capability of batteries. For example, a small electrode thickness, low electrode compaction density, and high electrode porosity can improve the transport of lithium-ions [23]. Low temperature significantly limits the ion conductivity and mobility of lithium-ion batteries [23]. However, there are relatively few studies on improving the low-temperature performance of lithiumion batteries in battery manufacturing. Therefore, this section attempts to summarize the existing methods that can improve the ion conductivity and mobility of lithium-ion batteries, in order to provide inspiration for improving the low-temperature performance of batteries in battery manufacturing in the future.
In this section, cell structure optimization based on tortuosity and porosity is reviewed and as shown in Figure 3, this figure shows the schematic diagram and result analysis of the improvement of lithium-ion battery performance by cell structural optimization. In a low-temperature environment, compared with a thick electrode, a thin electrode can greatly improve the conduction efficiency of lithium-ions, but has inferior energy storage potential. The electrode thickness ranges from a few tens to several hundred micrometers [24]. Zhou et al. [25] proposed the concept of the "sub-thick electrode". The sub-thick electrode refers to the electrode with a certain minimum thickness (300-2000 µm) directly manufactured by chemical reaction compared with the traditional thick electrode. Wang et al. [26] have also studied the ultra-thick double-layer LiFePO 4 (LFP) electrode with a thickness of up to 1200 µm. In the future high-power energy storage systems of electric vehicles, thicker electrodes will mean higher energy storage potential [27]. Data shows that, if the electrode thickness is increased from 25 µm to 200 µm, the proportion of inactive ingredients in the battery will be reduced from 44 wt% to 12 wt%, which means an increase in battery energy density [28]. Therefore, the development of thick electrodes with high lithium-ion conduction efficiency is promising for the applications of LIBs at low temperatures. Low electrode tortuosity and high electrode porosity improve the lithiumion transport in an electrode [28]. According to the equation D e f f = ε τ D (where ε is the porosity, D is the intrinsic ionic conductivity, and τ is the tortuosity), the effective ionic conductivity, D e f f , is inversely proportional to the electrode tortuosity, τ, and proportional to the electrode porosity, ε. Ju et al. [29] quantitatively established the relationship between the tortuosity of the ferric oxide electrode and the rate capacity of the battery through experiment and theory, as shown in Figure 3A. It can be found that, with the bending from 10 to 1, the rate capacity of the battery continues to improve. Thus, the electrodes with low-tortuosity structures exhibit fast ionic transport.   [29]. (B) Schematic diagrams of ionic transport paths in different electrodes. (a) Traditional electrodes with inherent zigzag pore structure. (b) Thick electrodes with low-tortuosity pore structure. The red arrows indicate the shortest ionic transport pathways [28]. (C) Full voltageareal capacity diagram and cycle performance of 3D-printed eight-layer LFP/LTO battery at different discharge rates [30]. (D) Schematics of functional EMP for adsorbing anions and facilitating lithium-ion transport [31]. Sun et al. [30] created a low-tortuosity electrode structure through 3D printing, which effectively improved the transport efficiency of lithium-ions. They uniformly dispersed Li4Ti5O12 (LTO) and LFP nanoparticles in a mixture of water, glycerin, ethylene glycol (EtG), and viscosifiers to produce a printable ink with a 60 wt% solid load. They used this ink for the 3D printing, followed by drying and annealing, of an eight-layer LTO/LFP  [29]. (B) Schematic diagrams of ionic transport paths in different electrodes. (a) Traditional electrodes with inherent zigzag pore structure. (b) Thick electrodes with low-tortuosity pore structure. The red arrows indicate the shortest ionic transport pathways [28]. (C) Full voltageareal capacity diagram and cycle performance of 3D-printed eight-layer LFP/LTO battery at different discharge rates [30]. (D) Schematics of functional EMP for adsorbing anions and facilitating lithiumion transport [31]. (E) Electrochemical characterization of PP and EMP. (a) Ionic conductivity at various temperatures and activation energy obtained from linear fitting of the Arrhenius equation. (b) Comparison of t Li+ and ionic conductivity [31].
The close, random packing of electrode elements in traditional thick electrode structures ( Figure 3B(a) [28]) usually results in electrodes with tortuous pore structures, which significantly hinders their penetration by electrolytes, considerably increases the ionic transport distance, and reduces ionic transport efficiency. As shown in Figure 3B(b) [28], Kuang et al. [28] developed a low-tortuosity pore structure to enhance the ionic transport in an electrode at low temperatures. The lithium-ion transport path in the innovative low-tortuosity electrode structure shortened, which effectively improved the lithium-ion transport efficiency.
Sun et al. [30] created a low-tortuosity electrode structure through 3D printing, which effectively improved the transport efficiency of lithium-ions. They uniformly dispersed Li 4 Ti 5 O 12 (LTO) and LFP nanoparticles in a mixture of water, glycerin, ethylene glycol (EtG), and viscosifiers to produce a printable ink with a 60 wt% solid load. They used this ink for the 3D printing, followed by drying and annealing, of an eight-layer LTO/LFP lithium-ion microbattery electrode structure with a low bending degree and a high aspect ratio. To study the electrochemical performance of the microbattery, Sun et al. [30] printed 8 and 16 layers of 3D-IMA (960 × 800 µm, 60 µm electrode width, 50 µm spacing) on a glass substrate and then measured the voltage-areal capacity curve of the microbattery at different discharge rates, as shown in Figure 3C(a) [30]. The areal capacity at 1 C was 1.5 mAh cm −2 , which agreed well with the theoretical results for LFP and LTO half-cells. Moreover, the cycle life of 3D-IMA was determined, and the results are shown in Figure 3C(b) [30]. The minimum attenuation of the capacity reached 30 cycles, showing a good cycle life.
High porosity can improve the lithium-ion transport in an electrode. High-porosity structures have been fabricated using different approaches, such as the subtractive design method, wherein a porous structure is introduced into the electrode by reducing some of the components of the electrode precursor [28]. During the drying process in traditional electrode manufacturing, by removing the solvent components in the electrode, an interconnected pore network is formed to improve the ion transport efficiency. However, this random electrode structure manufacturing approach increases tortuosity. Researchers have explored the orderly directional introduction of pore structures into electrodes to reduce tortuosity. For example, the transport efficiency of lithium-ions can also be improved using magnetic materials and external magnetic fields [28,32]. In such approaches, active electrode materials with anisotropic structures, such as graphite anodes, are modified and oriented [28,32] and then decorated with superparamagnetic ferric oxide (Fe 3 O 4 ) nanoparticles. To induce a sensitive magnetic response in such materials, Billaud et al. [33] applied an external magnetic field to the electrode slurry to control the orientation of graphite electrode sheets along the magnetic field and form an out-of-plane-aligned architecture. This unique structure decreased the lithium-ion transport distance and resulted in an ultrathick 10 mg cm −2 graphite electrode (200 µm thickness), which increased the specific charge capacity threefold.
Zhang et al. [31] used electrospinning to prepare a new composite separator called electrospun metal-organic framework (MOF)-poly (vinyl alcohol) composite membranes (EMP), which contains MOF particles and polyvinyl alcohol. The common polypropylene (PP) LIB separator was selected as the control group for experimental verification. They measured the Li + conductivity of the composite separator via electrochemical impedance spectroscopy. As seen in Figure 3D [31], at an ambient temperature of 30 • C, the ionic conductivities of 1 M LiPF 6 in the ethylene carbonate/diethyl carbonate LP's EMP (LP-EMP) and 1 M LiClO 4 in the propylene carbonate (LC)'s EMP (LC-EMP) were 2.9 and 1.9 mS cm −1 , respectively. In contrast, the ionic conductivities of LP-PP and LC-PP were only 0.7 mS cm −1 and 0.5 mS cm −1 , respectively. Besides, the activation energy of LP-EMP and LC-EMP were lower than the latter with PP. As shown in Figure 3E [31], the lithium-ion transference numbers t Li+ of LP-PP and LC-PP were 0.37 and 0.49, respectively, whereas those of LP-EMP and LC-EMP were significantly higher (0.59 and 0.79, respectively). In LP and LC, when PP was replaced with EMP, the conductivity of lithium-ions increased 467% and 433%, respectively, which significantly improved the battery conductivity. These findings indicated that MOF particles containing open metal sites could spontaneously adsorb anions and effectively transport lithium-ions in the electrolyte, thereby significantly increasing t Li+ (up to 0.79) and lithium-ion conductivity [31]. Therefore, the use of such MOF particles can provide new ideas for improving the performance of LIBs at low temperatures.
In general, cell fabrication significantly influences the low-temperature performance of batteries, but further study is needed to explore solutions and measures to improve conductivity. Through Table 1, it can be seen that, with the low tortuosity of the electrode, the development of a new EMP composite separator significantly improves the conduction efficiency and ion mobility of lithium-ions, increasing by 300% and 460%, respectively [31,33]. Besides, the directional introduction of pores also increases the capacitance by three-times. They proved that, in the future, we can take an important step in improving the rate performance and capacity performance of LIBs at low temperatures from the perspective of low-tortuosity electrodes and directional porous electrodes. The t Li+ is 460% higher than before [31] half-cells C/10 and 1 C, rate capability tests Increase the specific charge capacity by threefold [33]

Material Optimization
Studying and improving cell materials, such as anode, cathode, and electrolyte materials, can effectively improve the low-temperature performance of LIBs ( Figure 4) [20,23]. This section mainly introduces studies on enhancing such LIB performance by enhancing the cell materials.

Chemistry Test Results
Ref.
Cell fabrication/断 structure optimization The ionic conductivity is 300% higher than before [31] Li|electrolyte|Li cells Potentiostatic polarization at voltage bias of 20 mV 断Test: AC impedance and DC potentiostatic polarization measurements The tLi+ is 460% higher than before [31] half-cells C/10 and 1 C, rate capability tests Increase the specific charge capacity by threefold [33]

Material Optimization
Studying and improving cell materials, such as anode, cathode, and electrolyte materials, can effectively improve the low-temperature performance of LIBs ( Figure 4) [20,23]. This section mainly introduces studies on enhancing such LIB performance by enhancing the cell materials.

Anode Materials
Carbon materials. Graphite is the most-commonly used LIB anode material, but its performance is significantly affected by low temperatures. At sub-zero temperatures, the intercalation/de-intercalation kinetics of the graphite anode are very slow, resulting in a significant loss of battery capacity. Low temperatures inevitably lead to a rapid increase

Anode Materials
Carbon materials. Graphite is the most-commonly used LIB anode material, but its performance is significantly affected by low temperatures. At sub-zero temperatures, the intercalation/de-intercalation kinetics of the graphite anode are very slow, resulting in a significant loss of battery capacity. Low temperatures inevitably lead to a rapid increase in charge-transfer resistance [34][35][36]. According to U.S. ARL, when the temperature drops from −30 • C to −40 • C, the charge-transfer resistance of the graphite electrode battery increases by 2000-times [37]. In addition, graphite anodes are prone to lithium dendrites when charged at low temperatures, which may short-circuit the battery and pose a great challenge to battery safety [36]. Figure 5 illustrates the improvement effect on the lowtemperature performance of LIBs for the cases mentioned in this section. Prelithiation treatment can enhance the low-temperature properties of graphite. The prelithiation treatment can prevent the negative-electrode-form SEI film from consuming the lithiumions removed from the positive electrode, but obtain lithium-ions from other places, so as to ensure that the lithium-ions removed from the positive electrode will not be wasted in the formation of the negative SEI film [37]. Liu et al. [38] designed a Li 3 V 2 (PO 4 ) 3 (LVP)/hard carbon battery, and its initial charge was used for prelithiation. The battery showed a capacity retention rate of over 67% at −40 • C relative to that measured at room-temperature; this figure is far superior to that of traditional LIBs ( Figure 5A [38]). Kim et al. [39] prepared a composite composition of edge-plane-activated graphite and an a-Si nanolayer (SEAG) as a hybrid anode with good low-temperature properties, and its initial coulombic efficiency reached 93.8%. They measured the battery's capacity retention using a battery circulator (TOSCAT-3100, Toyo system). At a current density of 1.75 mA cm −2 , the capacity retention rate of this anode material in 50 cycles was 99.3% ( Figure 5B [39]). Nano-tin-doped particles can also improve the working ability of graphite at low temperatures [40]. Yan et al. [40] doped graphite with nano-tin as the anode (Sn/EtG) of an LIB. In their experiment, they found that the anode material decreased the charge-transfer resistance of the cell after 10 cycles at −20 • C. As shown in Figure 5C [40], the charge-transfer resistance of the anode material increased from 10 2 to 10 4 Ω as the temperature dropped from 25 • C to −20 • C, whereas the charge-transfer resistance of a common graphite anode was too large to be measured at −20 • C. Zhang et al. [41] developed a 3D nitrogen-doped porous carbon frame with a nitrogen content of 19% by sintering nitrogen-doped carbon dots at 800 • C. The specific capacity of the cell reached 840 mAh g −1 after 1000 cycles. The experiment was carried out with Land CT2001A at a current density of 2 Ag −1 from 0.01 to 3 V. Similar to silicon, nitrogen, and tin, Cu has been used to coat a graphite oxide anode [42]. The experiment in this previous study proved that this method could improve the battery capacity at low temperatures; however, the Cu coating may have compromised the cell capacity at 0 • C to 20 • C [42]. In short, researchers have made much exploration on how to improve the low-temperature performance of carbon material anode batteries, among which prelithiation treatment and anode surface modification are currently more-commonly used and -effective methods. Lithium titanate. The LTO structure is stable and enables lithium-ions to be deintercalated during (de-)lithiation [37]. The lattice change of LTO before and after phase transition is not obvious, which can be considered as zero strain, which is an important reason why LTO is considered to be a promising anode material [37,45,46,47,48]. However, the ion diffusion coefficient of LTO is not high (10 −9 to 10 −13 cm 2 s −1 ) and the conductivity is low (10 −9 S cm −1 ), which means that the low-temperature performance of LTO will be greatly limited [46]. The LTO particle size can be changed to reduce the interface resistance and enhance the properties of the cell [49]. Kallio et al. [49] explored the low-temperature performance of LTO particles with different sizes, including particles with a small primary particle size (LTO-SP, 150 nm, 22 m 2 g −1 surface area) and particles with a large primary particle size (LTO-LP, 225 nm, 7 m 2 g −1 surface area). They tested the charge/discharge capacity of LTO half cells at −20 °C and 0.1 C with the Neware battery cycler. The half cells were charged and discharged at constant current at a voltage range of 1.1 and 2.5 V, and they kept the temperature stable for an hour before the test began. The experiment verified that the capacity of the small particles (large surface area) was approximately 31% higher than that of the large particles. The reason may be that, at low temperatures, the smaller particle size will significantly increase the number of surface sites for lithium insertion and shorten the diffusion path of lithium-ions and may form a denser composite structure, which is very favorable for improving the capacity of the battery [49].
Silicon-based anodes. The theoretical capacity of silicon anodes is very high, especially at subzero temperatures; its capacity is much higher than that of graphite [37]. Therefore, silicon anodes are considered to be a method to improve the low-temperature performance of graphite anode batteries. Markevich et al. [43] used an Arbin model Lithium titanate. The LTO structure is stable and enables lithium-ions to be deintercalated during (de-)lithiation [37]. The lattice change of LTO before and after phase transition is not obvious, which can be considered as zero strain, which is an important reason why LTO is considered to be a promising anode material [37,[45][46][47][48]. However, the ion diffusion coefficient of LTO is not high (10 −9 to 10 −13 cm 2 s −1 ) and the conductivity is low (10 −9 S cm −1 ), which means that the low-temperature performance of LTO will be greatly limited [46]. The LTO particle size can be changed to reduce the interface resistance and enhance the properties of the cell [49]. Kallio et al. [49] explored the low-temperature performance of LTO particles with different sizes, including particles with a small primary particle size (LTO-SP, 150 nm, 22 m 2 g −1 surface area) and particles with a large primary particle size (LTO-LP, 225 nm, 7 m 2 g −1 surface area). They tested the charge/discharge capacity of LTO half cells at −20 • C and 0.1 C with the Neware battery cycler. The half cells were charged and discharged at constant current at a voltage range of 1.1 and 2.5 V, and they kept the temperature stable for an hour before the test began. The experiment verified that the capacity of the small particles (large surface area) was approximately 31% higher than that of the large particles. The reason may be that, at low temperatures, the smaller particle size will significantly increase the number of surface sites for lithium insertion and shorten the diffusion path of lithium-ions and may form a denser composite structure, which is very favorable for improving the capacity of the battery [49].
Silicon-based anodes. The theoretical capacity of silicon anodes is very high, especially at subzero temperatures; its capacity is much higher than that of graphite [37]. Therefore, silicon anodes are considered to be a method to improve the low-temperature performance of graphite anode batteries. Markevich et al. [43] used an Arbin model BT2000 battery tester and Autolab System to analyze the performance of a silicon anode in a 1 M LiPF 6 fluoroethylene carbonate (FEC)/dimethyl carbonate (DMC) (1:4 vol) solution at −30 • C. They found that the charging capacity of the silicon anode was 28.8-times that of the graphite anode. In addition, the team also measured the electrochemical impedance spectra (EIS) of two different anode cells at −30 • C with a potentiostat-galvanostat Model 128 N Autolab in the frequency range of 100 kHz-1.7 mHz [43]. As can be seen from Figure 5D [43], the resistance of the silicon anode was significantly lower than that of the graphite anode, which is one of the reasons for the superior capacity performance of silicon over graphite. Subburaj et al. [50] assembled a Si/LiNi 0.8 Co 0.15 Al 0.05 O 2 pouch cell and tested its electrochemical characteristics. At −40 • C, the discharge capacity of the cell exceeded 700 mAh g −1 , which was 65% of that at 20 • C. However, silicon-based anodes may have certain problems. For example, the volume expansion of silicon electrodes is severe, even reaching 300% [50]. Therefore, methods such as the use of nanostructures and porous structures have been implemented to enhance the practicality of silicon anodes at low temperatures [51,52].
In addition to the three common anode materials mentioned above, Sn, Ge, and other elements are often used as battery anodes [53][54][55]. SnO 2 has a natural advantage for low-temperature applications. Low temperatures inhibit Sn coarsening and maintain the high reversibility of the SnO 2 alloying and conversion reactions, thus ensuring stable capacity [44]. Tan et al. [44] experimentally compared the performance of a tin oxide anode and other anodes, such as graphite, at −10 • C. The electrolytes they used were 1 M LiFP 6 in a mixture of EC, PC, and EMC with a volume ratio of 1:1:2 and 5 wt% FEC additives. After 10 cycles, the capacity retention rate of the tin oxide electrode at −10 • C was 82.9% of that at 30 • C ( Figure 5E [44]). By contrast, the capacity retention rate of the graphite anode under the same conditions was only 10.1%, which was considerably lower than that of the tin oxide anode [44]. Table 2 summarizes the methods for improving the anode, including prelithiation treatment, particle size reduction, exploration of new anode materials, etc. Prelithiation treatment or exploration of new materials can significantly improve the performance rate and capacity of the battery. For example, at −40 • C, prelithiation treatment can maintain 67% capacity of the batteries that could not work [38], and the capacity of some new anode materials has been studied to be tens of times higher than that of graphite, such as Si [43] or SnO 2 [44]. However, the improvement of the low-temperature performance of the battery by reducing the particle size was not great, such as reducing the particle size of the LTO anode, which can only increase the capacity by up to 31% [49]. In fact, no matter what method, there are some limitations, which still need to be further explored. The capacity retention is 680% higher than that of graphite anode [44] Existing research indicates that carbon materials are the most-common anode materials. However, graphite and other such materials are significantly affected by low temperatures; in particular, the charge-transfer resistance rapidly increases with a decrease in temperature [34][35][36]. In addition, lithium plating is a major problem in carbon materials. Although mitigated via prelithiation, coating, and other methods, these problems still need further improvement [36]. Compared with carbon materials, lithium titanate is more stable and can, thus, enhance the low-temperature properties of LIBs [37]. However, the high working voltage and low energy density of this material limit its application to LIBs [51]. Silicon-based anodes exhibit good application potential at low temperatures [37]. Nevertheless, their severe volume expansion should be resolved, and studies on this subject are limited [49]. Hence, enhancing the low-temperature properties of LIBs by optimizing their anode materials has broad research prospects.

Cathode Materials
Cathodes may not be the most-critical factor resulting in the poor low-temperature properties of LIBs [20]; instead, they are essential parts of LIBs and are the core of Li + exchange. They also affect the working ability of cells at low temperatures, so cathode materials should be studied to improve the cryogenic properties of cells [20]. At present, reducing the particle sizes of materials [20,56], adding coatings [57,58], and doping elements [59] are important ways to enhance the low-temperature performance of batteries from the cathode perspective. Figure 6 shows the improvement in LIBs' performance at low temperatures after applying these methods.
Batteries 2023, 9, x FOR PEER REVIEW 12 of 31 elements [59] are important ways to enhance the low-temperature performance of batteries from the cathode perspective. Figure 6 shows the improvement in LIBs' performance at low temperatures after applying these methods. Layered oxides. Lithium cobalt oxide is a typical layered oxide, but its kinetic characteristics are poor. At −40 °C , the energy density of lithium cobalt oxide cathodes is only 5% of that at 20 °C [65]. In addition, NCM is a layered oxide similar to the structure of lithium cobalt oxide. During the charging process of the NCM cathode battery, the electrolyte is easily decomposed due to the instability of Ni 4+ , and the product will cover the NCM electrode, which will seriously affect the migration of Li + . In other words, the instability of Ni 4+ will lead to an increase in the internal resistance of the battery, which has a significant impact on the performance of the battery at low temperatures [66]. Tan et al. [67] coated NCM111 with glassy lithium borate (Li3BO3) sized approximately 8 nm. They used a charge-discharge apparatus to charge and discharge the battery at a rate of 0.2 C between 2.5 and 4.5 V to measure the battery's capacity. During cycling at 0.2 C and −40 °C , the specific capacity of the electrode coated with lithium borate was 65 mAh g −1 higher than that of an uncoated electrode. Moreover, the charging voltage of the lithium borate-coated battery was the same as that of the ordinary battery at the same capacity, but the former had a higher discharging voltage [67]. This was because the glassy lithium borate coating protected the cathode surface from the liquid electrolyte and unnecessary side reactions, thus reducing the polarization of the cell. Li3BO3 can also be deposited on Li1.2Ni0.2Mn0.6O2 particles to form coatings that have an appropriate thickness and can protect cathodes from HF attacks and ensure rapid Li + transport [67]. In addition, the team's analysis of EIS showed that the B-coating can greatly reduce the contact between the active substance and the electrolyte, thus reducing the charge-transfer resistance at Layered oxides. Lithium cobalt oxide is a typical layered oxide, but its kinetic characteristics are poor. At −40 • C, the energy density of lithium cobalt oxide cathodes is only 5% of that at 20 • C [65]. In addition, NCM is a layered oxide similar to the structure of lithium cobalt oxide. During the charging process of the NCM cathode battery, the electrolyte is easily decomposed due to the instability of Ni 4+ , and the product will cover the NCM electrode, which will seriously affect the migration of Li + . In other words, the instability of Ni 4+ will lead to an increase in the internal resistance of the battery, which has a significant impact on the performance of the battery at low temperatures [66]. Tan et al. [67] coated NCM111 with glassy lithium borate (Li 3 BO 3 ) sized approximately 8 nm. They used a charge-discharge apparatus to charge and discharge the battery at a rate of 0.2 C between 2.5 and 4.5 V to measure the battery's capacity. During cycling at 0.2 C and −40 • C, the specific capacity of the electrode coated with lithium borate was 65 mAh g −1 higher than that of an uncoated electrode. Moreover, the charging voltage of the lithium borate-coated battery was the same as that of the ordinary battery at the same capacity, but the former had a higher discharging voltage [67]. This was because the glassy lithium borate coating protected the cathode surface from the liquid electrolyte and unnecessary side reactions, thus reducing the polarization of the cell. Li 3 BO 3 can also be deposited on Li 1.2 Ni 0.2 Mn 0.6 O 2 particles to form coatings that have an appropriate thickness and can protect cathodes from HF attacks and ensure rapid Li + transport [67]. In addition, the team's analysis of EIS showed that the B-coating can greatly reduce the contact between the active substance and the electrolyte, thus reducing the charge-transfer resistance at low temperatures and increasing the ion diffusion coefficient, which is another reason why this method can improve the low-temperature performance of the battery [67]. Zhao et al. [60] coated Li 1.2 Ni 0.13 Co 0.13 Mn 0.54 O 2 with 2% aluminum fluoride. The batteries were charged and discharged between 2 and 8 V to test their capacity. As shown in Figure 6A [60], at −20 • C, the capacity of the cathode coated with aluminum fluoride was approximately two-times that of an uncoated cathode. However, at 25 • C, the coating only increased the discharge capacity by about 4.5%. Therefore, aluminum fluoride coating is an effective method to improve the performance of LIBs at low temperature. The spinel structure generated by the aluminum fluoride coating on the cathode material improved the electron migration, and the presence of LiAlO 2 on the coating sample enhanced the lithium diffusion kinetics, which may be the reason for the effectiveness of this method [60]. Moreover, decreasing the material particle size can enhance the characteristic of layered oxide cathodes at low temperatures [20,61]. Sun et al. [61] studied three LiNi 0.6 Co 0.2 Mn 0.2 O 2 (NCM622) materials with different nanostructures, particle sizes, pore distributions, and specific surface areas. Their synthesis process diagram is shown in Figure 6B [61]. The three modified cathodes enhanced the low-temperature performance of the cells, and the NCM622 material with the smallest particle size exhibited the most-remarkable improvement. The initial coulombic efficiency of the NCM622 material with a particle size of 3 µm reached 94.9%, and after 300 cycles at 0 • C, its capacity retention rate was 100% ( Figure 6C [61]). Element doping is also an effective way to enhance cell performance [59]. For example, Ti doping can change the lattice parameters of cathodes during charging and discharging and reduce impedance, thereby enhancing the low-temperature properties of cells [68]. Li et al. [68] charged and discharged the batteries between 2.8 and 4.25 V and tested its discharge capacity. At −20 • C, the discharge capacities of a Ti-doped NCM111 cathode were 51.3, 78.5, 101.8, and 122.4 mAh g −1 at 5, 1, 0.2, and 0.1 C, respectively, and the discharge capacities of a cathode without Ti were 118.8, 96.3, 67.7, and 30.1 mAh g −1 , respectively, under the same conditions.
Spinel-structured oxides. Spinel-structured oxides have excellent rate capability and physical stability. However, the charge-transfer resistance of this material increases exponentially with the decrease of temperature, and its low-temperature performance is much worse than that at room-temperature. [37,69]. Bi et al. [62] coated the structure surface of LiNi 0.5 Mn 1.5 O 4 (LNMO) with a continuous layer of Li 2 O-Al 2 O 3 -TiO 2 -P 2 O 5 (LATP), which improved the performance of LNMO. They made electrochemical measurements of the batteries between voltage windows of 3.5 and 5 V using the CT2001A battery test system. The result showed that the LATP-coated electrode (LATP-LNMO) helped reduce polarization at −20 • C, and its energy density at this temperature far exceeded that of an uncoated electrode (P-LNMO). The initial charge-discharge curves of the LATP-coated and uncoated cells are shown in Figure 6D [62]. Regardless of temperature, the discharge capacity of the coated electrode was high, and the electrochemical performance of the cell was remarkably improved by the LATP coating when the temperature was very low. The discharge capacity of the LATP-LNMO cathode was 82.24 mAh g −1 at −20 • C, which was considerably higher than that of the uncoated cathode (48.10 mAh g −1 ). In addition, the EIS showed that the charge-transfer resistance increased rapidly with decreasing temperature, but the charge-transfer resistance of the cells with LTAP coating was only about one-tenth of that of the uncoated cells at −20 • C (at 25 • C, the charge-transfer resistance of the two was of the same order of magnitude, with a difference of less than 30 Ω), indicating that the LTAP coating was favorable for charge-transfer at low temperatures [62]. Zhang et al. [63] doped the surface of LiMn 2 O 4 (LMO) with nickel, which reduced the charge-transfer barrier and inhibited the capacity loss caused by a drop in temperature. As shown in Figure 6E [63], the capacity retention rate of the Ni-doped LMO was significantly high at low temperatures (approximately 90% at −20 • C), whereas that of an undoped LMO was only approximately 65%.
Polyanionic compounds. LFP is a typical polyanionic compound with low cost and good thermal stability, but its conductivity (10 −9 S cm −1 ) and lithium-ion diffusion coefficient (10 −14 -10 −16 cm 2 S −1 ) are relatively low, resulting in the poor low-temperature performance and rate performance of LFP [20,[70][71][72][73][74][75]. Zhang et al. [64] added Mg and La ions into LFP, and the CT2001A battery test system was used to test the charging and discharging capacity of this cathode and a pure LFP cathode −20 • C and 10 C between 3.5 and 5 V voltage windows, and the measurement curve is shown in Figure 6F [64]. The cathode discharge capacity after La and Mg doping was 85.4 mAh g −1 , which was substantially higher than that of the pure LFP. At the same time, Zhang et al. [64] also used the CHI604B electrochemical station to perform the EIS. At −20 • C, the charge-transfer resistance of the pure LFP electrode was as high as 1431 Ω, while that of the doped LFP electrode was only 391.4 Ω. The conductivity of the pure LFP electrode was obviously improved, which can explain the reason why doping La and Mg elements can improve the performance of the LFP electrode [64]. Cui et al. [76] prepared a carbon-coated, phosphorusdoped LFP/C-P composite material as a battery cathode and tested the rate properties of a pristine commercial LFP cell, a pristine synthetic LFP cell, and the synthetic LFP/C-P composite at different temperatures. The experiment was carried out in a 2001T battery measurement system with a voltage range of 2.7-4.2 V. The experimental results showed that the rate performance of the LFP/C-P composite was excellent compared to that of the uncoated materials at any temperature. For example, at 0 • C and 1.5 C, the specific capacity of the battery containing LFP/C-P was 142.6 mAh g −1 , whereas that of the pristine commercial LFP battery was approximately 110 mAh g −1 . Moreover, at −25 • C, the specific capacity of the LFP/C-P cell was still approximately 40 mAh g −1 higher than that of the pristine commercial LFP cell [76]. C coating and P doping can alleviate the corrosion of the electrolyte on the cathode and can also establish a certain number of interconnecting channels, shorten the diffusion path of lithium-ions, and then, improve the capacity retention rate and rate performance of the battery at low temperatures [76]. In addition to the above methods, reducing the material particle size can accelerate the solid diffusion of LFP. For example, an LFP/C cell with a small particle size can shorten the transport paths of the ions, thus improving the solid diffusion kinetics [23]. However, this method may lead to side reactions [23,77]. LVP is also a polyanionic compound commonly used as a cell cathode material, and its theoretical capacity is higher than that of LFP [78,79]. Compared with LFP, LVP is more conducive to the extraction/intercalation of lithiumions [79], and coating it with carbon further improves its low-temperature performance [80]. Luo et al. [80] prepared an LVP cell coated with hierarchical carbon and verified that the coating improved the rate performance and cycle stability of the cell at −20 • C.
Vanadium oxides, Ni-based Prussian blue, and others can also be used as LIB cathode materials [81,82]. For example, V 2 O 5 has a layered structure and high lithium insertion potential [82]. Sides et al. [83] prepared a nanoscale (70 nm) V 2 O 50 . At −20 • C, the cathode material shortened the transport distance of Li + , thus enhancing the low-temperature performance.
Most studies on low-temperature cathodes focus on improving their lithium-ion transport performance, conductivity, and cathode stability [20,23]. Table 3 summarizes the experiments involving improved battery cathodes involved in this paper. As can be seen from the table, the common methods include doping, coating, and reducing the particle size. These methods can increase the discharge capacity by double or even four-times and improve the rate performance of the battery, but this will also bring some problems. For example, the addition of element coatings and the reduction of the particle size may produce by-products, which may lead to cell capacity reduction or pollution [51]. As we all know, doping and coating are not only for the improvement of low-temperature performance, but can optimize the electrochemical performance of the battery as a whole [57][58][59]. In order to conform to the theme, this paper mainly selected some cases with very significant improvement in low-temperature performance and briefly analyzed the mechanism. Although the examples mentioned in this article also slightly improved the performance of the battery at room-temperature, the improvement was not obvious compared to the low temperature, as shown in Table 4. The layered nanorods assembled with a small particle size have a stable structure, fast ion transport, a large surface area, full contact between the electrolyte and cathode, and good capacity reversibility The discharge capacity of the minimum particle size is 27.6% higher than that of the maximum particle size  The discharge capacity increased by 15.0% [76] From the discussion in this section, we can find that the improvement effect of various schemes on different cathode materials is different on the whole. For layered oxides and spinelstructured oxides, coating is a much more-effective method than doping and particle size reduction, while for polyanionic-type compounds, the opposite is true; doping elements are far more-effective than coating. In summary, the problem of enhancing the low-temperature properties of LIBs using their cathodes needs further study; different methods should be selected according to the cathode materials of different cathodes; new methods or materials should be explored.

Electrolytes
As seen in the Introduction, at low temperatures, high electrolyte viscosity, high charge-transfer resistance, and severe lithium plating on the anode considerably affect cell performance [20], so new electrolytes should be studied and designed to enable LIBs to work well at low temperatures. Figures 7 and 8 show the effect of optimizing the electrolyte on the low-temperature performance of the battery.
Aqueous electrolytes. The advantages of aqueous electrolytes include excellent safety, low cost, lack of environmental pollution, and high ionic conductivity [84,85]. However, pure water generally solidifies at 0 • C, which is unfavorable for battery use at low temperatures [37]. To overcome this defect, Ramanujapuram et al. [86] added 16 mol kg −1 LiCl, 9 mol kg −1 LiNO 3 , and 3.5 mol kg −1 Li 2 SO 4 to water to reduce the freezing point of the solution to −40 • C to 50 • C, −20 • C to 30 • C, and −30 • C to 35 • C, respectively. The capacity retention rate of the cell with LiCl at −40 • C was 72% of that at room-temperature; thus, LiCl is more dominant than conventional organic electrolytes. In addition to this, the researchers established electrochemical impedance spectros-copy (EIS) to study the effect of electrolyte resistance. Figure 8A [86] shows the EIS model used to fit the data. To eliminate the effect of external resistance, plot the Xintercept of the impedance curve as a function of distance, as shown in Figure 8E [86]. The slope of the line in the figure gives the resistivity of the electrolyte per unit length, and plot a function of temperature, as shown in Figure 8F [86]. The true electrolyte resistance is then obtained by repeating the test for each electrolyte at each temperature. Figure 8B-D [86] compares the electrolyte resistance, surface layer resistance, and charge-transfer resistance between −40 • C and 25 • C. The results showed that the increase of the three resistances was not obvious at −35 • C and the increase of the three resistances was sharp at −40 • C, while the increase of the charge-transfer resistance was the largest [86].

Electrolytes
As seen in the Introduction, at low temperatures, high electrolyte viscosity, high charge-transfer resistance, and severe lithium plating on the anode considerably affect cell performance [20], so new electrolytes should be studied and designed to enable LIBs to work well at low temperatures. Figures 7 and 8 show the effect of optimizing the electrolyte on the low-temperature performance of the battery.
Aqueous electrolytes. The advantages of aqueous electrolytes include excellent safety, low cost, lack of environmental pollution, and high ionic conductivity [84,85]. However, pure water generally solidifies at 0 °C, which is unfavorable for battery use at low temperatures [37]. To overcome this defect, Ramanujapuram et al. [86] added 16 mol kg −1 LiCl, 9 mol kg −1 LiNO3, and 3.5 mol kg −1 Li2SO4 to water to reduce the freezing point of the solution to −40 °C to 50 °C, −20 °C to 30 °C, and −30 °C to 35 °C , respectively. The capacity retention rate of the cell with LiCl at −40 °C was 72% of that at room-temperature; thus, LiCl is more dominant than conventional organic electrolytes. In addition to this, the researchers established electrochemical impedance spectros-copy (EIS) to study the effect of electrolyte resistance. Figure 8A [86] shows the EIS model used to fit the data. To eliminate the effect of external resistance, plot the X-intercept of the impedance curve as a function of distance, as shown in Figure 8E [86]. The slope of the line in the figure gives the resistivity of the electrolyte per unit length, and plot a function of temperature, as shown in Figure 8F [86]. The true electrolyte resistance is then obtained by repeating the test for each electrolyte at each temperature. Figure 8B-D [86] compares the electrolyte resistance, surface layer resistance, and charge-transfer resistance between −40 °C and 25 °C . The results showed that the increase of the three resistances was not obvious at −35 °C and the increase of the three resistances was sharp at −40 °C, while the increase of the charge-transfer resistance was the largest [86].  In addition, adding antifreeze additives to aqueous electrolytes can prevent solvent crystallization. EtG is widely used in the automobile, electronics, military, aviation, and many other fields because of its low freezing point. Figure 9 [87] shows the action mechanism of EtG in a 1 M Li 2 SO 4 aqueous electrolyte solution. Artur et al. [87] studied aqueous rechargeable lithium-ion batteries (ARLBs) with LFP cathodes without and with different amounts of EtG in 1 M Li 2 SO 4 aqueous electrolyte solutions. The results showed that the crystallization temperature of the electrolyte solution decreased from −4.6 • C to −24.6 • C with the addition of different amounts of EtG ( Figure 7A [87]), which greatly improved the low-temperature working range of the ARLBs [87].
In addition, adding antifreeze additives to aqueous electrolytes can prevent solve crystallization. EtG is widely used in the automobile, electronics, military, aviation, a many other fields because of its low freezing point. Figure 9 [87] shows the acti mechanism of EtG in a 1 M Li2SO4 aqueous electrolyte solution. Artur et al. [87] studi aqueous rechargeable lithium-ion batteries (ARLBs) with LFP cathodes without and w different amounts of EtG in 1 M Li2SO4 aqueous electrolyte solutions. The results show that the crystallization temperature of the electrolyte solution decreased from −4.6 °C −24.6 °C with the addition of different amounts of EtG ( Figure 7A [87]), which grea improved the low-temperature working range of the ARLBs [87].  [86]. (B-D) Comparison electrolytes, electrolyte surface layers, and charge-transfer resistance with temperature in aqueo and organic electrolyte systems (1 M LiPF6 in 1:1 EC: DEC) [86]. (E) Relation between X-intercept impedance curve and electrolyte distance [86]. (F) Comparison of resistivity and temperature three electrolytes [86]. (G) Influence of different electrolytes on voltage capacity of lithium-i batteries [91].  [86]. (B-D) Comparison of electrolytes, electrolyte surface layers, and charge-transfer resistance with temperature in aqueous and organic electrolyte systems (1 M LiPF 6 in 1:1 EC: DEC) [86]. (E) Relation between X-intercept of impedance curve and electrolyte distance [86]. (F) Comparison of resistivity and temperature of three electrolytes [86]. (G) Influence of different electrolytes on voltage capacity of lithium-ion batteries [91].
Solid electrolytes. All-solid-state cells can improve the safety of LIBs, but have poor capacity at low temperatures [20,92,93]. Lin et al. [94] designed a solid electrolyte made of γ-(2,3-epoxypropoxy) propytrimethoxysilane and boron (B) materials co-cross-linked with starch (B-starch-Si (BStSi)) [20,94]. The Li/LFP battery with BStSi electrolyte was placed in the refrigerator (Haier, BCD-220WDVL), and the battery's performance was tested in the battery test system (LAND CT2001A, BTRBTS) under an anti-static charge-discharge cycle at −20 • C. After 200 cycles at −20 • C, the capacity of the Li/LFP cell containing this electrolyte reached 55.9 mAh g −1 , which was 46.5% of its room-temperature capacity. This greatly enhanced the performance of the lithium battery at low temperatures and extended the working time of the battery [94]. Li et al. [95] designed and synthesized a covalent organic frameworks (COFs) solid electrolyte; after lithiation, the COFs were stripped to form covalent organic nanosheets, which promoted single-ion conduction without a solvent. Consequently, it quickly combined with and separated from the lithium-ions, thereby greatly enhancing ionic conductivity (10 −5 S/cm at −40 • C) [95]. Solid electrolytes. All-solid-state cells can improve the safety of LIBs, but have poor capacity at low temperatures [20,92,93]. Lin et al. [94] designed a solid electrolyte made of γ-(2,3-epoxypropoxy) propytrimethoxysilane and boron (B) materials co-cross-linked with starch (B-starch-Si (BStSi)) [20,94]. The Li/LFP battery with BStSi electrolyte was placed in the refrigerator (Haier, BCD-220WDVL), and the battery's performance was tested in the battery test system (LAND CT2001A, BTRBTS) under an anti-static chargedischarge cycle at −20 °C. After 200 cycles at −20 °C, the capacity of the Li/LFP cell containing this electrolyte reached 55.9 mAh g −1 , which was 46.5% of its room-temperature capacity. This greatly enhanced the performance of the lithium battery at low temperatures and extended the working time of the battery [94]. Li et al. [95] designed and synthesized a covalent organic frameworks (COFs) solid electrolyte; after lithiation, the COFs were stripped to form covalent organic nanosheets, which promoted single-ion conduction without a solvent. Consequently, it quickly combined with and separated from the lithium-ions, thereby greatly enhancing ionic conductivity (10 −5 S/cm at −40 °C) [95].
Cryogenic ionic liquid electrolytes. Ionic liquids have a wide working temperature range and can enhance the low-temperature characteristic of LIBs [20,96,97,98]. bis(trifluoromethanesulfonyl)imide (TFSI)-based ionic liquids exhibit good, stable electrochemical performance [20]. Aguilera et al. [99] introduced N-butyl-Nmethylpyrrolidinium (PYR14)-TFSI into a traditional electrolyte; its freezing point was lower than that of a carbonate-based electrolyte, and it enhanced the low-temperature performance of the cell. Ionic liquids are attracting extensive attention because of their wide electrochemical window and remarkable safety [100]. In the study of Balducci et al. [88], the ionic liquid N-butyl-N-methylpyrrolidiniumbis(trifluoromethanesulfonyl)imide (PYR14TFSI) and propylene carbonate (PC) were mixed in different proportions to form electrolytes with various conductivities at different temperatures. Conductivity is Cryogenic ionic liquid electrolytes. Ionic liquids have a wide working temperature range and can enhance the low-temperature characteristic of LIBs [20,[96][97][98]. bis(trifluoromethanesulfonyl)imide (TFSI)-based ionic liquids exhibit good, stable electrochemical performance [20]. Aguilera et al. [99] introduced N-butyl-N-methylpyrrolidinium (PYR14)-TFSI into a traditional electrolyte; its freezing point was lower than that of a carbonate-based electrolyte, and it enhanced the low-temperature performance of the cell. Ionic liquids are attracting extensive attention because of their wide electrochemical window and remarkable safety [100]. In the study of Balducci et al. [88], the ionic liquid N-butyl-N-methylpyrrolidiniumbis(trifluoromethanesulfonyl)imide (PYR14TFSI) and propylene carbonate (PC) were mixed in different proportions to form electrolytes with various conductivities at different temperatures. Conductivity is measured using a slave-phase-controlled Alpha analyzer and a slave-phase-controlled temperature controller. Figure 7B [88] shows that the conductivity of an electrolyte with 20% and 50% PYR14TFSI added at −30 • C to 0 • C was higher than that of an electrolyte without PYR14TFSI; moreover, the conductivity of the 20% group decreased more slowly with the decrease in temperature. It can be shown that the addition of 20% PYR14tfsi prevented the electrolyte from freezing, at least up to −40 • C [88]. Makoto et al. [89] used the ionic liquid 1-ethyl-3-methylimidazolium fluoride (EMIF)·2.3 HF as an electrolyte, and the conductivities of electrolytes were measured by a conductivity meter (Toa Electronics, CM-50S/CGT-511B). The results showed that the conductivity of their capacitor reached 20 mS cm −1 at −40 • C, which was better than that of a non-aqueous electrolyte (1 M Et 3 MeNBF 4 /PC) ( Figure 7C [89]). Although some ionic liquid electrolytes are challenging to obtain, they play an important role in the development of low-temperature LIBs.
Organic electrolytes. Organic electrolytes are the most-traditional LIB electrolytes. Solvents, lithium salts, and additives generally need to be adjusted to prepare organic electrolytes [101][102][103]. Therefore, enhancing the low-temperature properties of cells from the perspective of organic electrolytes should begin with these three aspects.
In terms of the solvent, low-melting-point cosolvents can be incorporated to improve the low-temperature properties of LIBs. Dong et al. [37,104] used a high-concentration electrolyte composed of 5 M LiTFSI, a dichloromethane (DCM) diluent, and an ethyl acetate (EA) electrolyte. The experimental data showed that the electrolyte had high ionic conductivity (0.6 mS cm −1 ) at −70 • C. As for lithium salts, and the anion modification of lithium salts may provide more lithium-ions to the electrolytes and improve ionic mobility [105]. Compared with traditional LiPF 6 -based carbonate electrolytes, LiBF 4 -, LiBOB-, and LiDFOB-based carbonate electrolytes can improve ionic mobility, thus enhancing the low-temperature working ability of LIBs [106]. In particular, the LiBF 4 -based electrolyte has a better working ability at low temperatures. Zhang et al. [91] prepared 1 M LiBF 6 and 1 M LiBF 4 electrolytes in a glovebox filled with argon gas and dissolved them in 1:1:3 PC/EC/EMC solvent to study the low-temperature performance of batteries with a graphite anode and lithium nickel-based mixed oxide cathode. Figure 8G [91] compares the cell polarization caused by the use of different electrolytes. As can be seen from the figure, the volt-capacity curves of LiBF 6 and LiBF 4 batteries were very close at 20 • C, but interestingly, the capacity of batteries containing LiBF 4 had a relatively significant increase at −30 • C, which indicates that LiBF 4 plays a good role in improving the lowtemperature performance of LIBs [91]. In addition to solvents and lithium salts, additives can remarkably enhance the low-temperature working ability of LIBs from the perspective of organic electrolytes. Additives can form stable SEIs, change the SEI composition, decrease the charge-transfer resistance, contribute to the uniform deposition of Li, and inhibit the growth of Li dendrites [105]; these additives include FEC, 1,3-propanesultane, and tris(trimethylsilyl) phosphor (TMSP). In addition, when lithium-silica sulfobetaine silane and hydroxyl-terminated poly(dimethylsiloxane) are mixed in a certain way, the synergistic effect between them can enhance the conductivity of the battery. This material can enhance the low-temperature properties of LIBs and is also a promising carbonate-based electrolyte additive [107]. Liu et al. [90] compared electrolytes E1 (0.05 M CsPF 6 containing 1 M LiPF 6 + PC/ethyl methyl carbonate (EMC)/ethylene carbonate (EC) [1:8:1 wt]), E2, and E3, which contained different additives ( Figure 7D [90]), to study the influence of the additive type and concentration on the low-temperature working ability of a GR/LiNi 0.08 Co 0.15 Al 0.05 O 2 (Gr/NCA) button cell. The cell with 0.5% of the electrolyte additive exhibited better discharge performance at −40 • C ( Figure 7D(a-e) [90]). The batteries containing E2 and E3 had better low-temperature performance, with the E2 electrolyte showing better performance ( Figure 7D(f) [90]). Gao et al. [108] designed a 1 M LiPF 6 + methyl acetate (MA)/EMC/EC/diethyl carbonate (DEC) (3:1:1:1 vol; baseline electrolyte (BE) + MA) electrolyte. They incorporated 1 wt% TMSP + 1 wt% 1,3-propanediolic sulfate (PCS) as an additive, and the resulting electrolyte was named BE + MA + TMSP + PCS. In the experiment, they used the LAND System (CT2001A, Wuhan, China) to charge and discharge LiNi 0.5 Mn 1.5 O 4 /MCMB full cells at different temperatures, and the current was calculated from the LiNi 0.5 Mn 1.5 O 4 cathode material (1 C rate set to 120 mA g −1 ). Then, they evaluated the low-temperature properties of a graphitic mesocarbon microbead (MCMB)/LNMO cell and concluded that the capacity retention rate of the cell with BE + MA + TMSP + PCS was 99.7% (101.7 mAh g −1 /102 mAh g −1 ) at a charge-discharge rate of 0.3 C at −5 • C. This showed that the battery using the BE + MA + TMSP + PCS electrolyte had better performance at low temperature [108]. The abovementioned FEC additives have been studied extensively. He et al. [109] tested the influence of FEC additives on the ionic conductivity of Li/lithium cobalt acid cells at −40 • C to 25 • C. At −40 • C, and the conductivities of electrolytes without FEC and with 5, 10, and 20 wt% FEC electrolytes were 5.42 × 10 −4 , 1.55 × 10 −3 , 2.35 × 10 −3 , and 1.63 × 10 −3 S cm −1 , respectively. At room-temperature, the ionic conductivities of electrolytes were 1.11 × 10 −2 , 8.55 × 10 −3 , 8.86 × 10 −3 , and 1.42 × 10 −2 S cm −1 , respectively. The electrolyte with 10 wt% FEC showed the maximum conductivity at both low and room-temperatures. The conductivities of the electrolytes without FEC decreased rapidly with a decrease in temperature. Adding an appropriate amount of the FEC additive (according to the experimental data, approximately 10 wt% FEC had the best effect) helped improve the ionic conductivity, and the lower the temperature, the more obvious the effect was. The researchers [109] also tested the effect of different concentrations of FEC additives on the battery discharge capacity. The experimental results showed that the capacity retention rate of the battery without the FEC additive was 66% and that of the cell having 10 wt% FEC was 77.1% at −40 • C and 0.2 C. The addition of FEC benefited the low-temperature performance of the battery. Their experimental results showed that the capacity retention rate of the battery with 10 wt% FEC was greater than not only that of the battery without the FEC additive, but also that of the battery with 20% FEC. Therefore, the FEC additive improved the low-temperature properties of the cells, but a higher FEC concentration did not lead to better battery performance [109]. In short, the design of electrolytes, including aqueous electrolytes, solid electrolytes, ionic liquid electrolytes, and organic electrolytes, has a considerable improvement in the discharge capacity of lithium-ion batteries at low temperatures and greatly extends the use time of batteries at low temperatures. Table 5 summarizes the experiments mentioned in this paper to improve the lowtemperature performance of LIBs from the perspective of the electrolyte. The data in the table show that improving the electrolyte is very helpful to enhance the ionic conductivity and can even increase the ionic conductivity by more than 60-times. The capacity retention rate increased by 16.8% [109] In general, researchers have used different methods to examine various types of electrolytes, but they all aim to improve the low-temperature properties of LIBs by reducing the freezing point of electrolytes, improving their ionic conductivity and improving their chemical properties, among others. The research process is hampered by many questions and difficulties, such as the selection of additives, the selection of additive volumes, and whether such added substances harm the environment. In addition, reducing additives may undergo oxidation reactions on the cathode, whereas FEC additives form thick SEI films on the graphite electrodes of PC and EC electrolyte systems [110]. All in all, the design and study of electrolytes at low temperatures need to be further pursued.

The Advantages and Disadvantages of Different Schemes
Each of the methods above has its advantages, but often comes with disadvantages. As shown in Table 6, for example, pre-lithiation treatment can improve battery capacity and energy density, but it cannot increase ionic conductivity. What is more, adding additives in electrolyte can enhance the electrochemical performance of the electrolyte at low temperatures, but some reducing additives undergo an oxidation reaction at the cathode, which is bad for LIBs. In order to obtain a better comprehensive performance of LIBs, it is suggested that material optimization should be combined with structure optimization. For example, a low-melting-point cosolvent can be added to the electrolyte at the same time as the anode material is prelithiated. This can not only inhibit the occurrence of lithium-plating in LIBs at low temperatures, but also improve the ionic conductivity.   [37,104] Lithium salt anion modification Improves ionic mobility May decrease electrolyte conductivity due to low solubility of some modified lithium salts in aprotic solvents [105]

Summary and Outlook
This review presented the current research status of the methods of enhancing the lowtemperature performance of LIBs at the cell design level. The low-temperature applications of LIBs here point to electric vehicles. First, the article highlighted the necessity of the development of new-energy vehicles and the benefits of LIBs as a power supply for new-energy vehicles from two current situations: the increasing demand for energy resources and the increasingly prominent environmental problems worldwide [1-3]. Low temperatures are then emphasized as an important factor limiting the use of new-energy vehicles. Next, this review introduced the deterioration of the low-temperature performance of LIBs and pointed out the main parameters to control or improve the performance of LIBs at low temperatures. Then, this review summarized four factors that deteriorate the performance of LIBs at low temperatures, which are the decreased ionic conductivity and wettability of liquid electrolytes, the increased intrinsic grain-boundary resistance and slow Li + diffusion rate in electrodes, and the difficult Li + dissolution and sluggish Li + transport through solid-electrolyte interphase (SEI) along with high charge-transfer resistance and the occurrence of Li plating [20], and explained what consequences these factors will lead to. Moreover, this paper illustrated the key consideration to improve the deteriorated capacity, so that readers can understand how to enhance the low-temperature performance of cells clearly. In addition, this paper summarized the methods to enhance the low-temperature performance of LIBs from two aspects and put forward the idea that the two should be combined.
Thus, several common methods of improving the low-temperature properties of cells were introduced in detail from the viewpoint of cell design. Methods to enhance the performance of the cell by optimizing the cell structure were introduced first. For example, electrodes with high porosity and low tortuosity can improve the transport efficiency of lithium-ions at low temperatures [23]. Electrode materials were then discussed. The effects of anode, cathode, and electrolyte materials on the low-temperature properties of LIBs were analyzed in detail. The common improvement methods of cathode and anode materials mainly include doping, coating, and particle size reduction. Coatings generally protect the cathode from corrosion, reduce polarization, and increase capacity, while doping elements can change the lattice parameters or chemical bonds of the cathode to slow the rate of increase of charge-transfer resistance at sub-zero temperatures. Reducing the particle size can increase the surface area of the electrode and shorten the diffusion path of lithiumions [56][57][58][59]. Here, the effects of these three methods on different anode and cathode materials were explained using examples, and the problems faced by these materials and methods, such as possible side reactions, were also highlighted. This review also discussed the most-common organic electrolytes, which can be improved by developing low-meltingpoint cosolvents, exploring new lithium salts, and studying various low-temperature additives. Moreover, the advantages of aqueous, solid, and ionic liquid electrolytes, along with their optimization methods for low-temperature applications were also discussed.
In general, from the perspective of cell design, the methods of improving the lowtemperature properties of LIBs include battery structure optimization, electrode optimization, electrolyte material optimization, etc. These can increase the reaction kinetics and the upper limit of the working capacity of cells. An electrolyte with a low freezing point, which can form a stable SEI with low impedance, needs to be developed. In addition, electrode optimization for high ionic conductivity, high energy density, and lithium plating prevention is a research hotspot [23]. Future studies can also focus on using various materials to induce synergistic effects for the improvement of cell materials.
Leaving aside the single point of view of battery design, future research should also combine battery management with battery materials to fully achieve the enhancement potential of the low-temperature properties of LIBs. The thermal management system can improve the working environment of the battery at low temperatures, such as air preheating [111], resistance preheating [112], phase change material preheating [113], self-heating techniques [114], and current excitation techniques [115]. Researchers have explored methods of enhancing the low-temperature properties of LIBs, but they only focus on a certain topic, and few scholars have proposed systematic solutions to the abovementioned problems. Combining material optimization schemes from the intrinsic aspect of batteries with thermal management schemes for an improved battery environment may result in a breakthrough in the low-temperature performance of batteries.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: