Zirui
Yang
,
Ruijuan
Shi
*,
Zhen
Shen
and
Yong
Zhao
*
Key Lab for Special Functional Materials of Ministry of Education; National & Local Joint Engineering Research Center for High-efficiency Display and Lighting Technology; School of Materials Science and Engineering; Collaborative Innovation Center of Nano Functional Materials and Applications; Henan University, Kaifeng, 475004, P. R. China. E-mail: 10330121@henu.edu.cn; zhaoyong@henu.edu.cn
First published on 14th June 2023
Na metal anode is one of most promising anode materials for next-generation secondary batteries. However, the practical application of Na anode is limited by dendritic growth, rapid volume change, and serious interface problems in the process of Na electroplating/stripping, resulting in low coulombic efficiency, short life, and safety issues of sodium metal batteries (SMBs). Herein, the cyclic instability mechanisms of the Na anode and the corresponding advanced protection strategies including in situ solid-electrolyte-interphase (SEI), artificial SEI, and three-dimensional conductive frame, are systematically reviewed. Notably, this review summarizes the latest research progress on interface modification and electrode modification of all-solid-state SMBs. Finally, the outlooks of anode interphase in SMBs are summarized and prospected, providing a promising way for high-energy and safe SMBs.
The ionic radius of Na+ (1.02 Å) is larger than Li+ (0.76 Å), which will affect the interphase substance, transport properties, and interphase formation. Compared with Li+, the larger Na+ with lower solvation energy undergoes a faster ion transport rate upon Na plating/stripping processes. Once Na is directly deposited on the anode during charging and discharging, rather than in the embedded compounds such as graphite stored in the sodium ion battery, the maximum gravimetric capacity of the battery will be significantly increased. In the field of room temperature SMBs, significant progress has been made in the cathodic side (such as layered oxide materials, polyphosphates, S, and O2).7–13 Two types of cathode materials have been developed for SMBs. The first type is intercalated materials such as anionic compounds, layered transition metal oxides, and Prussian blue analogs (PBAs). The intercalated cathode is mostly utilized in high voltage SMBs, although its specific capacity is constrained. The second type is conversion materials that include S, O2, metal oxides, and sulfides, which always have high specific capacities. For example, the theoretical energy density of the Na–S battery is up to 1274 W h kg−1 due to the high capacities of S cathode (1675 mA h g−1) and Na anode (1166 mA h g−1).14–17
Compared to the widely investigated cathode materials that also have been successfully applied to sodium ion batteries, the bottleneck on the negative side of SMBs should be paid more attention. Generally, Na anode is slightly irreversible during the repeated plating and stripping due to its high reactivity with the electrolyte. During the process of Na deposition and stripping, Na with high reductive activity easily reacts with electrolyte and forms a solid electrolyte interphase (SEI) on its surface. Notably, the original SEI layer generally experiences the uneven flow of ions and low transfer rate of Na+ on the Na surface, resulting in the fracture and recombination of SEI films, as well as electrolyte depletion during battery operation.18–29 Thus, the uneven deposition and stripping of Na surface are easy to cause the growth of Na dendrites and side reactions, leading to low coulombic efficiency (CE), poor stability, and low safety for SMBs. The above issues faced by the Na anode largely limit its practical process in large-scale energy storage systems (Fig. 1). In the past several decades, significant progress has been made in the development of high-performance SMBs. The structural/interfacial engineering on Na metal anode is emerging to regulate Na plating/stripping behavior and improving the cycling stability, while few reviews focus on the interface issues and advanced strategies both in liquid electrolyte and solid-state electrolyte (SSE) of SMBs.30–34 Moreover, few reviews have related to causes of the key mechanisms behind various issues for Na metal and how to handle electrode/electrolyte interactions in SSEs, which could provide advice and guidance on how to build stable SEI layers in solid-state SMBs.
Fig. 1 Primary problems and corresponding solutions of Na anodes.30,31,151,193,198 (Reprinted with permission from ref. 30, 31, 151, 193 and 198. Copyright 2017 American Chemical Society, Copyright 2018 WILEY-VCH, Copyright 2017 WILEY-VCH, Copyright 2016 WILEY-VCH, and Copyright 2018 WILEY-VCH). |
Herein, we have highlighted the failure mechanisms of the chemical reactions between Na anodes and electrolyte during battery operation, along with advanced strategies (using SEI layer modification and 3D conductive frameworks) to address potential issues about electrode/electrolyte interfaces of SMBs in liquid and SSEs. In this review, we primarily discussed the instability mechanism of Na electrode during cell cycling, which is mainly due to the side reaction and dendrite growth on the surface of Na metal. According to the failure mechanism, this review has elaborated the Na protection strategies from three aspects: in situ SEI layer, artificial (ex situ) SEI layer, and three-dimensional (3D) frame structure. Notably, the advanced optimization strategies of electrolyte-electrode interface and electrolyte structures in SSE systems are discussed to deal with the poor interface compatibility and slow ion transport problems of SSEs. The electrochemical and chemical behavior of SMBs changes greatly in complex charge/discharge process, which will be reflected in this review. Furthermore, the prospects and challenges of developing high-stable Na anodes are comprehensively presented as well, which provides a promising way to achieve high-energy and safe SMBs in varied applications.
Fig. 2 (a) Schematic relative electron energies map of electrodes and the electrolyte in a stable battery.35 (b) For the circulation of Li/Li and Na/Na symmetrical batteries in LP30 at 25 °C, 1 M NaPF6 is added in EC0.5DMC0.5 or 1 M NaPF6 is added in EC0.45PC0.45DMC0.1, and the current density was 5 mA cm−2.36 (Reprinted with permission from ref. 35 and 36. Copyright 2009 American Chemical Society and Copyright 2015 Electrochemical Society). |
The side reactions between Na and traditional liquid electrolytes are inevitable without SEI modification, even worse in ester-based electrolytes. For example, Ponrouch's research showed that the main components of the SEI layer contained inorganic components (Na2CO3) and organic components (RCOONa and HCOONa) in carbonate ester-based electrolytes.36 As shown in Fig. 2b, the Li/Li symmetrical cell with LiPF6 ethylene carbonate (EC)/dimethyl carbonate (DMC) electrolyte showed fairly steady plating/stripping behavior at a current density of 5 mA cm−2. In contrast, the Na/Na symmetrical cell in the NaPF6 EC/DMC electrolyte exhibited extremely unstable cycling performance, demonstrating poor interface compatibility between Na metal and carbonate ester-based electrolytes. Moreover, impedance testing revealed the continuous reactions between Na and ester-based electrolyte, resulting in a gradual rise in the interface resistance of Na/Na symmetrical cell under open circuit conditions.
Considering the high reactivity of Na metal, it is easy to react with active intermediates in the electrolyte to produce several side reactions, which is usually accompanied by gas evolution as well. For instance, Mullins’ team reported that the deposition of Na metal in ester-based electrolytes (the hybrid of EC, diethyl carbonate (DEC), and PC) can produce a large amount of gas and fragile porous dendrites (Fig. 3a–f).37,38 The addition of fluoroethylene carbonate (FEC) solvent in DEC significantly reduced gas emissions during Na deposition, demonstrating that the formation of a dense SEI layer could prevent further decomposition of the FEC/DEC (1:1 volume) electrolyte and improve the cycling stability of Na metal anode. The rigid-shelled button batteries are insensitive to gas generation, while it is crucial to research the gas generation in pouch-type cells. Zhang et al. reported that the ion-solvent complexes with lower LUMOs (vs. pure PC solvent) in the NaClO4/PC electrolyte were prone to react with Na metal, leading to serious electrolyte depletion and gas evolution (Fig. 3g–i).39 Once PC complexed with Na+, the LOMO level of the resultant Na+–PC complex was reduced to −5.28 eV, accompanied by the intense gas generation on the surface of Na metal. Except for the serious side reactions between Na anode and ester-based electrolytes, Cui and colleagues uncovered that ether-based electrolytes with NaPF6 could enable Na anode with highly reversible and nondendritic stripping/plating behavior owing to the impermeable and even inorganic SEI made of NaF and Na2O.53 However, ether-based electrolytes electrochemical with narrow electrochemical windows cannot endure long-term exposure to high-voltage and high-temperature testing, causing undesirable electrolyte depletion and evolution of flammable gas on Na anode under extreme conditions.199 Moreover, ionic liquids and high concentration electrolytes have both been explored in addition to the conventional ester and ether electrolytes used in SMBs, providing steady high voltage performance and a sizable suppression of Na dendrites.
Fig. 3 All images corresponding to Na matrix before and after deposition in the (a and b) EC/DEC, (c and d) PC/FEC, and (e and f) FEC/DEC/DEC electrolytes. (g) Frontier molecular orbital theory analysis.37 The frontier molecular orbital level of PC (single PC molecule), PC + Sol (PC molecule considering solvent effect), [NaPC] (Na atom PC complex), and [NaPC]+ (Na+ ion PC complex). H, Li, C, and O atoms were marked in white, purple, gray, and red, respectively. Schematic diagram of orbital hybridization between PC and (h) Na ion and (i) Na atom.39 (Reprinted with permission from ref. 37 and 39. Copyright 2014 American Chemical Society and Copyright 2018 WILEY-VCH). |
To sum up, highly reactive Na metal is prone to lose the outermost electrons and then react with the electrolyte, thus affecting the Na deposition and evolution behavior and cycling stability of SMBs. The generation mechanism of side reactions is not solitary, and it is closely linked to unstable SEI and Na dendritic growth. The formation of porous Na dendrites increases the contact area between the Na anode surface and the electrolyte, which intensifies the side reactions and gas generation problems. As a result, the gas evolution process must be addressed from the standpoint of the overall SMBs as well as must cope with unstable SEI and Na dendritic formation problems.
Generally, the sequential growth process is used to explain how Na dendrites grow, which demonstrates how the uneven SEI film is caused by the inhomogeneity of the Na surface. The electroplating of Na+ on the inhomogeneous Na surface is not uniform, causing ion transport to concentrate in the bumps, and eventually form dendrites.45,46 The formation of Na dendrites induces the fracture and recombination of SEI films, electrolyte depletion, and safety issues for SMBs. Firstly, the separator is pierced by the Na dendrite formation, which could lead to an internal short circuit of SMBs. Meanwhile, the battery would become extremely hot, which could result in significant safety incidents. Secondly, some Na dendrites are still present. A portion of the dendritic will be discharged into the electrolyte when the stripping position is near the base of the dendrite, producing “dead Na”, leading to low CE and capacity decline of SMBs. In addition, “dead Na” may cling to the surface of Na, causing the polarization of the battery to gradually rise and eventually fail.38 Therefore, Na dendrites formation will affect the cycle stability and safety of SMBs, and it is crucial to induce the uniform Na+ flux distribution and Na stripping/plating process on the surface of Na anode.
Compared with the traditional SIBs, the Na+ storage in the SMBs is largely from the Na metal anode. The metal anode used in the SMBs is a process of electroplating through Na+ to convert into solid metal deposition. In this process, the size difference between Na atom and Na+ is an important reason for the volume change of Na anodes.47,48 The volume change of solid metal is significantly higher than that of traditional graphite intercalation materials. Traditional intercalation materials will provide a specific space to accommodate Na+, while the electrochemical process of sodium metal electrode takes place on the anode surface. The anode suffers from the pulverization and volume change problems during the cyclic deposition/stripping process, with “dead sodium” scattered in the electrolyte (together with the growth process of Na dendrite). This uncontrollable volume change in the Na anode will eventually increase the over-point position on anode surface, increase the internal resistance, and damage the cycle stability of SMBs.49
The original SEI layer faces difficulties due to the volume change of the sodium metal anode. The SEI layer will crack when the mechanical strength of the produced SEI is insufficient. The sodium metal electrode is continuously corroded by the electrolyte from the location where the SEI layer cracks, resulting in the loss of Na metal and electrolyte, which would decrease the CE and cycle life of SMBs.42 In general, the volume expansion of Na metal anode could be accommodated by 3D conductive frames, which will be discussed in-depth in the coming chapters.
Fig. 4 Development history of interface treatment methods for Na anode.30,31,60–64 (Reprinted with permission from ref. 60, 61, 30, 31, 62, 63 and 64. Copyright 2015 American Chemical Society, Copyright 2017 The author(s), Copyright 2017 American Chemical Society, Copyright 2018 WILEY-VCH, Copyright 2020 WILEY-VCH, Copyright 2022 Published by Elsevier B.V. and Copyright 2022 The Royal Society of Chemistry). |
Strategies for Na metal protection | Advantages | Disadvantages |
---|---|---|
Artificial SEI | 1. The SEI layer with high mechanical strength can be constructed | The SEI layer ruptures and it cannot be repaired |
2. Na electrodes can be treated with substances insoluble in conventional electrolytes | ||
3. The SEI layer with complex components can be constructed | ||
In situ SEI | 1. In situ SEI layer can be formed continuously in the presence of electrolyte additive | 1. When the effective substance in the electrolyte is depleted, it cannot be repaired to ensure the integrity of the SEI layer |
2. The in situ construction is simple for battery operation | 2. Additives should be soluble in defined electrolytes | |
3. The composition of the in situ SEI layer is the individual component | ||
3D skeleton | 1. 3D skeleton can reduce the local current density on the Na anode surface | 1. 3D skeleton may consume a large amount of electrolyte |
2. 3D skeleton with a porous structure can accommodate the volume change during Na striping/plating processes | 2. The 3D skeleton may decrease the whole mass energy density of SMBs |
(1) In situ SEI construction.50–52 To promote Na+ transport and address battery polarization brought by slow ion transport rate, the SEI film can be produced in situ on the surface of Na metal electrode using electrolyte additives. In addition, the Na electrode can be effectively protected using the chemical inertness of the in situ SEI film, and the low CE initiated by the chemical reactions between the electrolyte and Na metal can be improved. The dendritic growth issue caused by surface inhomogeneity can also be resolved by the production of a uniform SEI layer.
(2) Artificial SEI modification.53–55 The artificial SEI layer is applied to block the direct contact between the Na metal and the electrolyte before cell assembly, which is convenient to adjust the mechanical strength and homogeneity of the SEI layer. Through the artificial construction of SEI film, it can effectively promote Na+ transfer rate and protect the Na metal anode from electrolyte corrosion.
(3) 3D conductive framework.56–59 The 3D conductive framework can reduce the current density of the electrode, promote the uniform deposition of Na, and reduce the formation of Na dendrite. Moreover, the 3D conductive frame can also accommodate Na metal, reducing the battery failure caused by the volume expansion of Na electrode during the deposition process.
As ether electrolytes come into touch with a sodium metal surface, they react to form a thick SEI layer with low permeability of Na+. Therefore, increasing interfacial ion conductivity of the SEI layer is the main subject in ether-based electrolytes. Ether-based electrolytes are widely used in LIBs, while the related studies on fluorine ether in SMBs are scarce.75–81 Yi et al. used an ether-based electrolyte consisting of 1 M NaPF6 in DME, FEC, and 1,1,1,3,3,3-hexafluoroisopropylmethyl ether (HFPM) (2:1:2, volume ratio) in SMBs.82 The addition of HFPM contributed to the formation of a stable SEI layer on the Na anode, reducing the vigorous reaction between FEC and Na metal. As a result, the Na||Na symmetric battery with the NaPF6-DME-FEC-HFPM electrolyte exhibited a cycle life of over 850 h and a low polarization voltage of ∼70 mV at 1 mA cm−2. In addition, the Na||NVP full cell with the NaPF6-DME-FEC-HFPM electrolyte showed good long-term stability after 2000 cycles and high average CE of 99.9% even at 5 C.
The original SEI layer formed in the ester electrolyte is loose and unstable due to the serious side reaction between abundant polar groups of ester solvent and Na metal. Therefore, the development of a uniform and thick SEI layer on the surface of Na metal anode is encouraged in ester-based electrolytes in SMBs. Significant advancements have been made in the study of lithium metal batteries because LiF is one of the primary elements of the SEI barrier.83,84 Similarly, NaF could be used to enhance the stability of SEI layers on Na anode by fluoridating the solvents in SMBs.
Solvent, sodium anions, and additives compete for space on the surface of the Na electrode, and the component with the strongest interaction with the electrode surface becomes the primary constituent of the SEI layer. Energy analysis suggests that materials with lower LUMO level will be reduced more frequently at a higher potential. However, the solvation effect in the solution will change the LUMO energy level of the substance. For instance, pure EC molecule has a higher LUMO level than that of the FEC molecule, while the solvation effect of EC with Na+ will enable EC with a comparable LUMO energy level to FEC in the electrolyte.39,85 Many researches have shown that the degree of fluorination of solvent will affect its solvation ability in electrolyte. The higher the degree of fluorination of the solvent, the lower the solvation ability in the electrolyte.86 As a result, the solubility of Na salt in the electrolyte will decrease under a high degree of fluorination of the solvent. Amine et al.87 discovered that FEC molecules cannot form a stable SEI layer on the Na surface until the solvation number is more than one. Therefore, it is important to balance the relationship between solvation ability and the degree of fluorination during the fluorination process of solvents. Lee et al. demonstrated a combined electrolyte of 1 M NaFSI and FEC solvent to enhance the reversibility of Na stripping/plating in Na/Cu cells, displaying outstanding cycling performance at 2.82 mA cm−2 and comparatively stable voltage response.88 In contrast, the Na||Cu cell in 1 M NaFSI in EC/PC (1/1) electrolyte with 1 wt% FEC showed a sharp capacity reduction at 0.056 mA cm−2, indicating the rapid consummation of 1 wt% FEC additive.
ILs dissolved with various Na salts could be used as electrolytes in Na-based batteries.96–98 By altering their structural makeup, ILs can give electrolytes better electrochemical and interface stability. Imidazonium, pyrrolionium, and phosphonium cations are common cations, and [BF4]−, [FSI]−, and [PF6]− are the corresponding anions,121 which will affect the properties of ILs with different combinations of ions. The bulk viscosity of the ILs rises along with the ionic radius of the cations, which leads to a decrease in the ionic conductivity. Generally, anions with large size do not always cause increased viscosity, while anions with asymmetric structure can affect the viscosity of ILs by altering its melting point.90
Typically, many of the anions in ILs are F-containing anionic groups, which contribute to the formation of the SEI layer on the surface of the Na anode. Besides, the F-containing anionic groups could combine with the added sodium salts, increasing the NaF component in the SEI layer.
Makhlooghiazad et al. used a tributylmethylphosphonium bis(fluorosulfonyl)imide (P111i4FSI) ionic liquid as an electrolyte and added different concentrations of NaTFSI to it for research.99 An Na||Na symmetric battery was assembled after adding 90% P111i4FSI/NaFSI electrolyte. The battery assembled using this electrolyte exhibits a polarization voltage of 10 mV at 90 °C (Fig. 5a). An impedance test was performed on it. The interface resistance after the cycle is significantly lower than the interface resistance before the cycle. This fully demonstrates that the formation of a layer of NaF on the Na metal surface is conducive to the transmission of Na+ and reduces the interface impedance of the battery. Through the energy dispersive X-ray spectroscopy (EDX), it can be clearly seen that there are a large number of P and F elements on the surface of Na metal, which can effectively protect the Na metal electrode and provide a faster ion transfer rate (Fig. 5b).
Fig. 5 (a) Galvanostatic curves of 90 mol% P1i444FSI/NaFSI at 90 °C at a current density of 0.1 mA cm−2 with a cycling period of 10 min. (b) EDX spectrum of 90 mol% NaFSI/P1i444FSI, showing high intensity of O, F, and S and low intensity of C and P.99 (c) Chemical structures of electrolyte components and the corresponding dipole moments of different cationic salts in ILs.100 (d) F 1s spectra of various anionic ions in ILs.100 (e) Schematic diagram of different anionic salts in ILs.101 (f) Nyquist plot recorded for the cells after 5 cycles.101 (g) The voltage profiles of the Na/Na symmetrical cells at 1.0 mA cm−2/1.0 mA h cm−2. (Reprinted with permission from ref. 99, 100 and 101. Copyright 2017 The Royal Society of Chemistry, Copyright 2022 American Chemical Society and Copyright 2019 American Chemical Society). |
In the ILs without fluorine, the effective component of the SEI layer on the surface of Na metal is affected by the solvation effect of cathode additives on the ILs. Forsyth et al.100 added fluorinated anions (PF6−, TFSI−, (fluorosulfonyl)(trifluoromethanesulfonyl)imide (FTFSI−), and FSI−) in a methylpropylpyrrolidinium dicyanamide ([C3mpyr]DCA) IL to regulate the interfacial electrochemistry of the SEI layer on the Na anode (Fig. 5c). On the surface of the Na electrode, NaFSI was promoted to form a stable SEI layer with NaF as its primary component (Fig. 5d), while bis(trifluoromethylsulphonyl)imide sodium salt (NaTFSI) and NaFTFSI mostly existed as fluorinated anions on the surface of Na metal.
Cationic regulation is also essential in ILs, and the polarity and redox durability of cations could influence the electrochemical performance of SMBs. Rakov et al. regulated the surface structure of Na metal with different polar cations in NaFSI-concentrated IL electrolytes (Fig. 5e).101 In comparison to the IL cations with more polar and greater dipole moment (μ), such as N-methyl-N-propylpyrrolionium [C3mpyr]+ and N-methyl-N-methoxy-methylpyrrolionium [C2O1mpyr]+, the less polar cations with small μ(N-methyl-N-ethylpyrrolionium [C2mpyr]+) exhibited tighter and more presence on the negatively charged electrode/electrolyte interface. The [C2mpyr]+ with the smallest μ was prone to accumulated copiously on the negatively charged surface of the Na anode, resulting in “blocking effect” with greater energy barrier for NaxFSIy during the Na deposition process. In contrast, the impedance experiments revealed that the NaxFSIy clusters in highly polar [C2O1mpyr]+ and [C3mpyr]+ systems prior to reduction were favorable for the generation of inorganic rich SEI (Fig. 5f). Moreover, it was found that the electrolyte composed of [C2mpyr]+ showed a higher polarization voltage than [C2O1mpyr]+ and [C3mpyr]+ through the symmetrical battery test, and the [C3mpyr]+ cations with relative reductive stability displayed the most stable SEI layer (Fig. 5g).
Name | M w [g mol−1] | T m [°C] | σ [mS cm−1] |
---|---|---|---|
NaClO4 | 122.4 | 468 | 6.4 |
NaBF4 | 109.8 | 384 | — |
NaPF4 | 167.9 | 300 | 7.98 |
NaOTf | 172.1 | 248 | — |
NaTFSI | 303.1 | 257 | 6.2 |
NaFSI | 203.3 | 118 | — |
The NaF-based in situ SEI layer could be constructed by modifying the structure of fluorine-containing Na salts, depositing a NaF-rich SEI layer on the surface of Na metal. Ma and colleagues introduced the 4-acetylpyridine (4-APD) with electron-withdrawing acetyl group to the NaPF6/EC/DEC electrolyte in SMBs. The acetyl group increased more PF6− anions coordinated in the Na+ solvation structure, accelerating the decomposition of PF6− to the form a NaF-rich SEI layer. With the 1.0 wt% addition of 4-APD, the Na/Na symmetric battery delivered a long cycling life for over 360 h at 1.0 mA cm−1.107 Besides, Cui et al. proved that NaF can protect the Na anode by adding NaPF6 into the ether-based electrolyte, forming an SEI film composited of Na2O and NaF on the surface (Fig. 6a).60 The primary element in organic components was RCH2Na. When different sodium salts (e.g., NaTFSI, NaFSI, NaOTf, and NaClO4) were added in diglyme, the CE of SMBs decreased obviously upon battery cycling (Fig. 6b).
Fig. 6 (a) The schematic diagram of the SEI layer formed on the Na metal surface using NaPF6 in the electrolyte. (b) The CE comparison of SMBs with various electrolyte salts in diglyme.60 (c) The proposed formation mechanism of SEI layer on Na anode in NaBH4/diglyme electrolyte. (d) The cycle performance of Na||Na symmetric cells in the electrolytes of (black) NaClO4 EC/PC + 5 wt% FEC, (red) NaOTf/TEGDME, (blue) NaBH4/diglyme, and (green) NaBH4/TEGDME.108 (e) Convenient preparation process of Na1 by the deprotonation of [HNMe3]+[HCB11H11]− with NaH.109 (Reprinted with permission from ref. 60, 108 and 109. Copyright 2015 American Chemical Society, Copyright 2022 Royal Society of Chemistry and Copyright 2022 Wiley-VCH GmbH). |
The existing protection methods of sodium metal have not addressed the effect of the F source in the SEI layer on Na metal, but the difference in the F source in lithium metal batteries is involved in the distribution of LiF and the effect on the cycle performance of Li metal, which is crucial to further studies for the cycle stability of SMBs.84
The addition of sodium salt containing F atoms into SMBs is thoroughly explained, which produces an NaF layer on the Na surface to stabilize the Na metal. However, F-containing salts have the tendency to hydrolyze in the presence of trace water and generate a certain amount of HF byproducts during the reaction process, which somewhat constrains their application in electrolytes. Another way to increase the interfacial stability of Na metal is building a F-free SEI layer on the surface of Na metal through the control of sodium salt anions without F.
Kim et al.108 created a “fluorine-free” electrolyte using NaBH4 as an electrolyte additive in diglyme in Fig. 6c, which produced the primary components of NaBO2 and NaH in the SEI layer on the Na surface, according to time-of-flight secondary ion mass spectrometry (TOF-SIMS). Under 1 mA cm−2 and 1 mA h cm−2 conditions, the assembled Na||Na symmetry battery with NaBH4/diglyme electrolyte showed that the polarization voltage was approximately 75 mV at 100 cycles and 60 mV at 500 cycles. In contrast, the symmetrical battery built by the comparison electrolyte of NaClO4 EC/PC with 5 wt% FEC failed after only 150 cycles (Fig. 6d). Tomich et al. tried the Na–Al half battery using fluorine-free [Na]+[HCB11H11]− (recorded as Na+1) carbon-aryl electrolyte (Fig. 6e).109 The average CE of the Na–Al half battery using Na+1/diglyme electrolyte was 98.6%, while the distributed CE for the NaPF6/diglyme electrolyte was 103%. As a result, this test made it clear that Na+1 in diglyme solution showed better electrochemical performance than that of NaPF6.
In addition to the modification of the SEI layer with different kinds of Na salt on the surface of Na metal, the electrode can be protected by changing the concentration of Na salt. By adding a large amount of sodium salts into the electrolytes, the high concentration electrolytes are achieved to produce a positive effect on the SMBs. High concentration electrolytes have two benefits on the solid–liquid interphase: (1) it can remove unbound solvent molecules from the Na surface, preventing Na metal corrosion and the formation of an unstable SEI layer. (2) There is a significant amount of enriched Na+ on the Na surface, which can both effectively decrease the Na dendrite growth on the Na surface as well as the battery polarization generated by the concentration difference. Chen et al. elaborated that Nax(FSI)y was concentrated on the anode surface to separate the organic solvent from the Na anode.135 By efficiently preventing the production of the unstable SEI film, this high concentration electrolyte might shield the Na anode from organic electrolyte corrosion. In addition, surface-attached FSI− could interact with Na to create NaF-based SEI layer, which promotes the buildup of Na+ and prevents Na dendrite growth. However, the Na+ transmission rate in the high concentration electrolytes decreased obviously due to their high viscosity. To cope with this problem, bis(2,2,2-trifluoroethyl)ether (BTFE) was added to form a locally-concentrated electrolyte (labelled as NaFSI/1,2-dimethoxyethane (DME)-BTFE). The CE of the assembled Na–Cu half-cells with different electrolytes was tested at 1 mA cm−2 and 1 mA h cm−2. The 5.2 M NaFSI/DME electrolyte was used to obtain an average CE of 98.5%, while the locally concentrated electrolyte of 2.1 M NaFSI/DME-BTFE (1:2) displayed an average CE of 98.95% after over 400 cycles. In Na||NVP, the 1.7 M NaFSI/DME electrolyte quickly failed. With a capacity retention rate of 90.8%, the 2.1 M NaFSI/DME-BTFE (1:2) electrolyte could deliver a capacity of 66.4 mA h g−1 at 20 C after 40000 cycles. In contrast, the Na||NVP cell with 5.2 M NaFSI/DME electrolyte exhibited a capacity of 24.0 mA h g−1 and a capacity preservation rate of 48%. Besides, many researches focused on solving the problem of low bulk-phase ionic conductivity of high concentration electrolyte,133–135 and the details would not be depicted in this topic.
Fig. 7 (a) Diagrammatic representation of the effect of CTAB electrolyte additive on the Na deposition procedure. (b) X-ray photoelectron spectroscopy depth profile analysis of the SEI layer produced in an electrolyte comprising CTAB and NaPF6-diethylene glycol.137 (c) Schematic diagram of the SMB with an all-fluoride electrolyte by the connection of a bridge molecule. (d) Comparison of constant current plating/stripping process of Na in Na/Na symmetric cells at 1.0 mA cm−2 and 1.0 mA h cm−2.138 (Reprinted with permission from ref. 137 and 138. Copyright 2022 American Chemical Society and Copyright 2020 The Royal Society of Chemistry). |
Organic electrolytes with FEC might create an SEI layer containing NaF on Na metal. Besides, FEC could be employed as an electrolyte potentially that retards flames. Luo et al. designed a safe SMBs based on 1.0 M NaPF6 electrolyte in FEC/PC/1,1,2,2,2-tetrafluoroethylene-2,2,3,3-tetrafluoropropylene ether (HFE) and perfluoro-2-methyl-3-pentanone (PFMP).138 In this electrolyte formulation, FEC acted as a “bridging solvent” to promote the mutual dissolution of the two liquids (Fig. 7c). In addition, FEC could produce NaF on the surface of Na metal to shield the Na anode from electrolyte corrosion and extend the battery life. It was clear from the precise voltage distribution that the conventional 1.0 M NaPF6 in ethylene carbonate (EC)/PC (1:1 vol%) (referred to as N-EP) electrolyte was experiencing significant voltage variations and an increasing overpotential (the 50th cycle was 760 mV). While utilizing 1.0 M NaPF6 in FEC/PC/HFE (referred to as N-FEPH) and N-FEPH + PFMP electrolytes, the plating/stripping polarization of batteries was restricted to 160 and 90 mV (Fig. 7d). Vinyl carbonate, trimethylsilane phosphite, Na difluoro(oxalic acid)borate, and other substances could be employed as organic electrolyte additives as well. The initial capacity of the Na/Na3V2(PO4)2O2F full cell evaluated at 0.5 C was 120 mA h g−1, and the CE was 94.7% with the N-FEPH + P electrolyte, compared to 89.1% with the N-EP electrolyte.
Jiang et al. suggested a method to create ULCE + 2 wt% BSTF by adding N,O-bis(trimethylsilyl) trifluoroacetamide (BSTF) as an adjuvant to 0.3 M NaPF6 + EC/PC (1:1 vol%) extremely low concentration electrolyte (ULCE) as the neutral electrolyte (blank).139 The introduction of BSTF served as an ionic carrier NaF, which possibly contributed to the removal of HF and H2O from the electrolyte and the inhibition of NaPF6 hydrolysis. In an impedance measurement, the interface impedance of ULCE + 2 wt% BSTF-modified electrolyte was 85.2 after 20 cycles on the Na electrode surface, which was lower than that of ULCE (461.4). The entire Na||NVP battery with ULCE + 2 wt% BSTF demonstrated the average CE of 92.63%, while the ULCE/blank-based cells (with 1 M NaPF6 EC/PC (1:1 vol%)) failed after 650 and 800 cycles, respectively. These results demonstrated that the BSTF additive can effectively promote the application of a low concentration electrolyte.
Moreover, the production of a functional organic polymer on the surface of Na metal can result in an in situ SEI layer as well. Wei et al. generated a ionic polymer film on the surface of Na anode through the electropolymerization of an imidazole cationic ionic liquid monomer (Fig. 8a).140 The surface impedance in the symmetrical battery using electrolyte without 1,3-diallyl imidazolium perchlorate (DAIM) grew unsteadily. By contrast, it was discovered that the impedance of the symmetrical battery increased consistently after 2 h of circulation with a 20 wt% DAIM electrolyte, demonstrating that an SEI layer developed over time on the electrode surface (Fig. 8b). The electrode impedance remained consistent, following a stable increase when DAIM was added into the battery, indicating that a stable SEI film had developed on the surface (Fig. 8c). As a result, the full battery showed a CE of 96% and a capacity of 97 mA h g−1 after 160 cycles (Fig. 8d).
Fig. 8 (a) Diagrammatic representation of the electropolymerization process of imidazole IL on Na metal surface. (b) Nyquist plots of Na||Na symmetric cell following every 2 h charging with DAIM at 1 mA cm−2. (c) SEM images of the Na surface in the symmetric coin cells with and without the addition of DAIM in the electrolytes after 10 rounds of stripping and plating (scale bar: 20 μm). (d) The cycle performance of the full battery with imidazole IL added into the electrolyte.140 (Reprinted with permission from ref. 140, Copyright 2017 WILEY-VCH). |
Carrier ion diffusion in the alloy anode typically results in compressive stress in front of the propagation contact, which prevents carrier ion diffusion and restricts the amount of carrier ions that can fully penetrate the alloy anode during the battery cycle. Self-limited diffusion (SLD) is a phenomenon that restricts the full usage of the anode materials and lowers the rate performance of the battery. Byeon et al. showed that the diffusion of Na into Sn crystal would cause tensile stress near the interface and promote the formation of high-density dislocations.141 By creating a pathway for dislocation tube diffusion and relieving the SLD effect, the ensuing dislocations facilitated the diffusion of Na+ at an ultrafast rate. Sn or Na–Sn alloy production can efficiently promote the deposition of Na+ and minimize dendrite formation on the Na surface. Huang et al. modified 1.0 M NaClO4 in EC/PC electrolytes using SnCl2 as an electrolyte additive (Fig. 9a and b).142 The surface of Na would react with the changed electrolyte to generate a Na–Sn alloy and a NaCl mixed SEI layer, protecting Na from reacting with the electrolyte by further passivating the Na surface. The symmetrical battery could cycle steadily for more than 500 h at an overpotential less than 50 mV after adding 50 mM SnCl2 into the carbonate electrolyte. The polarization of the blank battery steadily increased to exceed 100 mV after about 200 cycles. The full Na||NVP battery with 50 mM SnCl2 additive showed a capacity of 87% at 1 C, while the blank one only retained 75% of its capacity. Xiang et al. added SbF3 into the electrolyte to generate NaSb and NaF layers on the surface of Na. The main component of the electrolyte was 2 M NaFSI/DME with 1,1,2,2-tetrafluoroethyl-2,2,3,3-tetrafluoropropyl ether (FEPE) and 1 wt% SbF3.143 The assembled full battery (with NVP as cathode) with SnF3 showed a capacity of 107 mA h g−1 at 1 C. The SEI film created after SnF3 modification could successfully promote the Na+ transmission rate and the capacity of the sodium metal battery at high magnification.
Fig. 9 The schematic diagram of (a) mosaic SEI and (b) dendrite-free morphology of the NaSn alloy layer on Na metal anodes. (c) Cycling stability of Na/Na symmetric cells with different SnCl2 amounts.142 (d) Schematic diagram of the Na plating process in the FEC and LiDFBOP electrolyte. (e) The cycle performance of Na||Na battery in E/P (1/1), F/D (1/1), and F/D (1/1) with 0.5% LiDFBOP electrolyte formula at a current density of 1 mA cm−2 and a capacity of 1 mA h cm−2.144 (f) The cycle performance of Na||Na3V2(PO4)2F3 cells in different electrolytes. (g) The galvanostatic plating and stripping behavior of Na/Na symmetrical cells at 5 mA cm−2 and 2.0 mA h cm−2 in the NaTFSI/TMP and NaTFSI/TMP-FEC-HFE electrolytes. (h) Cycling stability of Na||NVP full cells in NaPF6/EC-PC and NaTFSI/TMP-FEC-HFE electrolytes.149 (Reprinted with permission from ref. 142, 144 and 149. Copyright 2019 American Chemical Society and Copyright 2022 Royal Society of Chemistry). |
The use of a salt composed of nonreducible ions (e.g., Li+ and K+) as a cationic additive is also a solution, which has functions including electrostatic shielding, optimization of electrolyte solvation structure, and electrolyte stabilization. Chen et al. used lithium difluorobis(oxalato)phosphate (LiDFBOP) as an additive in 1 M NaPF6 FEC/DEC (v:v = 1:1, labeled F/D(1/1)) electrolyte in SMBs (Fig. 9d).144 DFBOP− anions induced a stable film through redox reaction with Na metal, and Li+ cations could the dendrite-free Na plating via electrostatic shielding, both promoting the stability of electrode/electrolyte interphases. Testing revealed that the symmetric batteries with treated electrolyte could maintain a steady cycle for 2600 h while maintaining polarization below 0.25 V (Fig. 9e) (surface capacity of 1 mA h cm−2, current density of 1 mA cm−2), but symmetric batteries without LiDFBOP failed after 880 h. In comparison to F/D (1/1) and EC/PC (1/1) electrolytes, the F/D (1/1) electrolyte with LiDFBOP demonstrated outstanding electrochemical performance (500 cycles of stabilization) in Na||NVPF throughout the entire battery test at 1 C (Fig. 9f).
SMBs can also receive the addition of K+ to stabilize the SEI layer on the Na anode. Ji et al. added KFSI into the 1 M NaPF6-EC/PC (NaPF-K) electrolyte to improve the cycling stability of SMBs.145 K+ served as an electron shield ion in the electrolyte to bind on the surface of Na metal. In addition, the P–F bond in PF6− was more readily broken than the S–F bond in FSI−, which caused a NaF/KF-rich SEI layer on the surface of Na metal. Following the battery test, the NaPF-K assembled symmetrical battery was able to run steadily for 400 h at a current density of 0.5 mA cm−2, and the polarization voltage increased from 260 to 305 mV. Therefore, the integrity of the Na anode during deposition and stripping could undoubtedly be ensured by the NaPF-K electrolyte.
Fullerene (C60) was discovered to have the ability to enhance the surface of Na metal, induce Na+ deposition after enrichment, and inhibit the growth of sodium dendrites, improving the long cycle life of SMBs. Li et al. bound –NO2 to the surface of C60 to create the C60(NO2)6 group, which was introduced as an additive to 1 M NaPF6/DME electrolyte at a concentration of 0.5 mM.146 The Na||Na symmetric battery with C60(NO2)6 electrolyte exhibited more than 500 cycles (1000 h) at 2.0 mA cm−2 and a low polarization voltage of less than 30 mV, while the unmodified blank electrolyte failed within 100 cycles. Moreover, the assembled Na||Cu battery with 1 M NaPF6/DME and C60(NO2)6 still maintained high CE even under the condition of 5 mA cm−2 and 5 mA h cm−2. Moreover, Li et al. formed C60(CF3)6 by combining C60 with the –CF3 group and adding it to 1 M NaPF6/DME electrolyte.147 The Na electrode surface was able to maintain stable Na+ transmission and reduce the polarization under the protection of C60 and the NaF-rich SEI layer. The symmetric battery in the electrolyte of C60(CF3)6 and 1 M NaPF6/DME demonstrated a cycle life for over 300 h and a polarization voltage of 30 mV at 5 mA cm−2 and 2 mA h cm−2, which was superior to the blank electrolyte. With an initial capacity of 108 mA h g−1 in Na||NVP@C full battery with C60(CF3)6 and 1 M NaPF6/DME electrolyte, 99% capacity retention could be achieved after 1000 cycles at 1.5 mA cm−2 but only 65% with 1 M NaPF6/DME under the same condition. It was evident that C60-based salts are promising in the electrolyte modification of SMBs.
Liu et al. introduced 1,1,2,2-tetra-fluoroethyl 2,2,3,3-tetrafluoropropyl ether (HFE) and FEC into 1 M NaTFSI trimethyl phosphate (TMP)-based electrolyte, which was assembled into a new high concentration electrolyte (1 M NaTFSI + HFE + FEC + TMP).149 With a high concentration solution, the symmetric cell made up of 1 M NaTFSI + HFE + FEC + TMP electrolyte maintained a polarization voltage of 120 mV after 1000 cycles at 0.5 mA cm−2 (Fig. 9h). Besides, the full Na||NVP battery with 1 M NaTFSI + HFE + FEC + TMP electrolyte showed a starting capacity of 100 mA h g−1 and a capacity retention of 93.1% after 400 cycles (Fig. 9g). In contrast, only one experiment could be conducted in the Na||NVP battery with the 1 M NaTFSI + TMP electrolyte. As a result, adding HFE and FEC could not only form a stable SEI on the surface of Na metal but also lessen the amount of free solvent molecules. According to the aforementioned results, sodium salts and other electrolyte additives are crucial for achieving long cycle and low polarization of SMBs, and the in situ SEI construction is a practical strategy to achieve high-CE SMBs. The recent reported artificial SEI and related approaches for protecting Na anodes are shown in Table 3.
Na@SEI | Electrolyte | Current density (mA cm−2) | Energy density (mA h cm−2) | Na/Na cell (cycle life@overpotential (mV)) | Ref. |
---|---|---|---|---|---|
Na@NaF | 1 M NaPF6 in DME + FEC + HFPM (2:1:2, in volume ratio) | 1 | 1 | 800 h@52 | 82 |
Na@NaF | 1 M NaPF6 + [C3mpyr]DCA | 0.1 | 0.1 | 50 h@10 | 100 |
Na@NaH + NaBO2 | NaBH4 + diglyme | 1 | 1 | 1200 h@50 | 108 |
Na@Na2CO3 + Na2O | Na+1 + DME | 2 | 1 | 300 h@30 | 109 |
Na@NaF | 2.1 M NaFSI/DME-BTFE (1:2) | 1 | 1 | 1000 h@>60 | 135 |
Na@NaF + NaBr + CTA+ | 5 mM CTAB + NaPF6-diglyme | 10 | 30 | 400 h@— | 137 |
Na@NaF | 1.0 M NaPF6 + FEC + PC + HFE + PFMP | 0.5 | 1 | 4400 h@90 | 138 |
Na@NaF | 0.3 M NaPF6 + EC/PC (1:1 vol%) + 2 wt% BCTF | 0.5 | 0.5 | 350 h@153 | 139 |
Na@DAIM | 20 wt% DAIM + carbonate-based electrolyte | 1 | 1 | 242 h@55 | 140 |
Na@Na-Sn + NaCl | 50 mM SnCl2 + 1 M NaClO4 + EC/PC | 0.5 | 1 | 500 h@100 | 142 |
Na@NaF + NaSb | 1 wt% SbF3 + 2 M NaFSI + DME + FEPE | 0.5 | 0.5 | 1200 h@950 | 143 |
Na@Li− + NaF | LiDFBOP + 1 M NaPF6 in FEC/DEC | 1 | 1 | 2000 h@150 | 144 |
Fig. 10 (a) Schematic diagram of the Li+/Na+ deposition process on stacked graphene on copper foam and comparison with the deposition overpotential of bare copper for Li+.150 (b) Schematic illustration of Na deposition behavior on the Na/NSCNT and bare Na electrodes.31 (c) Flow chart of Al2O3 thin film preparation on Na surface using ALD technology. (d) The stripping/plating curves of symmetrical batteries with bare Na electrode and the electrode after ALD treatment.152 (e) SEM images before and after Na deposition on NSCNT-modified and unmodified Na electrodes. (Reprinted with permission from ref. 150, 31 and 152. Copyright 2018 WILEY-VCH, Copyright 2018 Published by Elsevier B.V. and Copyright 2016 WILEY-VCH). |
In addition, carbon-based materials with flaws or doped heteroatom structure is better suited to enhance Na+ transport and further foster Na+ deposition. Wang et al. used N and S-treated carbon nanotubes (NSCNTs) as the interlayer to control the nucleation behavior of Na and inhibit the formation of Na dendrites (Fig. 10b).31 The sodium affinity of functional groups containing N and S elements on carbon nanotubes could induce the uniform distribution of Na on the NSCNTs layer. The SEM images revealed that the treated anode surface was flattened after 20 cycles (Fig. 10e). The initial polarization voltage of the Na/NSCNT||Na/NSCNT battery was about 70 mV, and the stable cycle could maintain 500 h at 1 mA cm−2. However, the unmodified Na||Na battery showed a lower initial polarization voltage (20 mV) but a short-circuit at 30 h.
The Na anode could reduce Sn4+ to produce the Na–Sn alloy and corresponding Na salts when touched with the solid or electrolyte of Sn salt. Kumar et al. reported a method of treating Na anode using gas–solid two-phase. The reaction between SnCl4 vapor and Na metal produced a surface layer containing NaxSny and NaCl (Fig. 11a).153 This two-material surface layer could effectively shield the Na from electrolyte degradation and guarantee the Na+ transfer rate throughout the Na deposition process. The reaction equation is shown as follows.
4Na + SnCl4 → Sn + 4NaCl, |
xNa + ySn → NaxSny. |
Fig. 11 (a) Graphical representation of the formation process of the artificial MAI on the Na metal interface. (b) Cross-sectional view of bare Na and MAI-Na interface. (c) The EIS of Na//Na cells without and with MAI after different stripping/plating cycles.153 (Reprinted with permission from ref. 153. Copyright 2020 Elsevier B.V.). |
Field emission scanning electron microscopy (FESEM) was used to describe the artificial metal alloy interphase (MAI), and the thickness of the MAI was approximately 6 μm (Fig. 11b). The particles were uniformly distributed across the surface of Na, and Na+ nearly diffused along the x-axis in this arrangement. The Nyquist curve revealed that the resistance of the MAI was about 3.5 Ω after the first cycle, while the resistance of the original Na anode was substantially greater. The resistance reduced further after 10 cycles, indicating the low Na+ diffusion barrier of MAI (Fig. 11c). The MAI-Na||MAI-Na cell showed stable cycles of over 650 cycles at a current density of 2 mA cm−2, while the untreated electrode showed a cycle life of less than 100 cycles. The comparison of the Na–S full cell (0.5 C) showed that the Na||S cell died after 50 cycles, but the average CE of the MAI-Na||S cell remained at 99.7%.
Na–Bi alloy as a protective layer can effectively prevent electrolyte corrosion of the Na anode and can provide excellent ionic conductivity. Sun et al. reported that Na powder and Bi powder were mechanically mixed and then compressed to prepare the mixed electrode (Fig. 12b).154 The mixed electrode could experience a spontaneous alloying event to increase metal surface tension and Na+ affinity, which could well alter the Na deposition and effectively limit dendrite formation. The symmetrical battery with mixed anode exhibited a overpotential of 20 mV after cycling for 1200 h in carbonate electrolyte, while the original Na electrode failed within 100 h with a polarization voltage of 120 mV. In a flame-retardant electrolyte, the electrode was employed. The full battery with Na3V2(PO4)3 cathode and treated Na showed a CE of 99.9% at 585 mA g−1 in a flame-retardant phosphate electrolyte. The Na–Bi alloy combined with Na salt could obtain a high-performance SEI layer as well. Yu et al. reacted Na with BiOCl to produce a mixed SEI layer containing Na3OCl and Na–Bi alloy (Fig. 12c).155 This SEI served two primary purposes: (1) promotes Na+ deposition and obtains consistent Na/artificial interface deposition. (2) Prevents the formation of Na dendrites. The SEI layer with Na3OCl possessed a high Young's modulus and interface energy. The symmetrical battery with treated Na anode demonstrated a low polarization voltage of 30 mV and a steady cycle of 700 h (in carbonate electrolyte) at 1 mA cm−2 (Fig. 12d). NVP||Na3OCl/Na3Bi@Na batteries showed excellent cycling performance at 15 C (1C = 110 mA g−1) and a reversible capacity of 83 mA h g−1 with a CE of ∼99.5% after 1500 cycles. In contrast, NVP||Na batteries exhibited rapid capacity attenuation, especially after 500 cycles, which could be attributed to the slow kinetics of the Na anode/electrolyte interface and “dead sodium” production. Chen et al. dropped 40 μL 10 mmol L−1 of Bi(SO3CF3)3 dimethylether electrolyte on the Na surface for the in situ reaction (Fig. 12e and f).156 The high reduction of Na resulted in the formation of an Na–Bi alloy, and the electrode resistance dropped considerably following treatment. The Na/Bi–O2 battery exhibited a discharge capacity of 1000 mA h g−1 over a cycle life of 50 cycles with a CE of 96.3%.
Fig. 12 (a) Manufacturing diagram of the Na-MoS2 (CR) hybrid. It consists of three steps: (1) dispersing massive MoS2 onto the surface of a piece of fresh Na, (2) folding the MoS2 Na hybrid, and (3) rolling the MoS2 Na hybrid.157 (b) Schematic diagram of the manufacturing process of the Na–Bi alloy foil. (c) The synthesis process of Na3OCl/Na3Bi@Na.157 (d) The cyclic performance of symmetric batteries with bare Na, Na3Bi@Na, and Na3OCl/Na3Bi@Na at 1.0 mA cm−2 and 1 mA h cm−2 (the inset was the schematic diagram of the Na3OCl/Na3Bi@Na anode).155 (e and f) Schematic diagram of Na deposition model on the Na and Na/Bi composite anode.156 (Reprinted with permission from ref. 157, 155 and 156. Copyright 2017 American Chemical Society, Copyright 2022 American Chemical Society and 2019 Wiley-VCH). |
Similar to the usage of Na–Sn and Na–Bi alloy as the main material for SEI modification, the use of Na–Mo alloy has also been studied. Yang et al. used MoS2 nanosheet to cover the surface of Na metal at room temperature to generate a composite SEI layer with Na2S, MoS2 and Mo (Fig. 12a).157 The Na2S created by the reaction could be widely disseminated on the surface of Na and used as a synthetic SEI film, inhibiting Na dendrite formation. The leftover MoS2 nanoslice could form a 3D frame layer on the Na surface, efficiently accommodating the volume expansion throughout the Na deposition process. The symmetric cell with the composite SEI layer showed a low polarization voltage of 10 mV, which was much lower than that with the bare Na (31 mV) or the Na2S-coated Na (90 mV) anodes.
2Na + 2H2O → 2NaOH + H2 |
2Na + 2NH3 → 2NaNH2 + H2 |
Fig. 13 (a and b) The schematic diagram of the surface on the Na anode showing the inhibition of large dendrite formation (due to the hardness interface of NaOH) and the regulation of large volume expansion during Na growth (due to the toughness of the NaNH2 interface). The stress and strain responses of NaOH (c) and NaNH2 (d) are calculated via DFT, respectively. The stripping/plating performance of symmetrical battery configurations with and without BPI at the current densities of (e) 40 mA cm−2 and (f) 50 mA cm−2.158 (f) Cycling stability of the symmetrical cell with/without the Na bromide layer. (g) Schematic diagram of the NaBr coating procedure. (h) The relationship between the surface binding energies of NaF (left) and NaBr (right) and the binding sites of adsorption atom diffusion by the joint density functional theory. (i) The voltage distribution current density obtained by continuously charging and discharging the symmetrical Na battery at 0.5 mA cm−2.61 (Reprinted with permission from ref. 61 and 158. Copyright 2020 The Author(s) and Copyright 2017 The Author(s)). |
The FESEM image revealed that the produced BPI (approximately 3 mm thick) could well cover the Na surface owing to the high stiffness and high ductility of the combined NaOH and NaNH3. As shown in Fig. 13c and d, NaOH showed a strong Young's modulus of 31.0 GPa, while the NaNH3 demonstrated stronger mechanical toughness, leading to a low Na+ diffusion barrier during the sodium plating and stripping process. As a result, the Na/Na symmetric cell exhibited stable cycling performance for 500 cycles even at a high current density of 50 mA cm−2 (Fig. 13e and f). In comparison with the sharp death of the Na||S cell without BPI within 100 cycles, the BPI-Na||S cell showed an average CE of about 98.3% for over 500 cycles.
NaBr and NaCl, which have similar structure to NaF, are also widely concerned by researchers as components of artificial SEI membranes. Archer's research group reported that the low surface diffusion barrier of NaBr in the SEI layer was favorable for rapid Na+ transport, resulting in the enhanced stability of the sodium deposition behavior.61 The NaBr-based protective layer was formed via the chemical reaction of Na metal with bromopropane after evaporation (Fig. 13g). Compared to the diffusion barrier of NaF and other lithium salts, the diffusion barrier for NaBr was only about 0.02 eV per atom due to the large anion size, which was calculated using joint density functional theory (Fig. 13h). The symmetric sodium cell with NaBr-coated Na metal showed stable Na plating/stripping behavior at 0.5 mA cm−2 (Fig. 13i). The CE of the NaBr-coated Na||S cell remained almost unchanged for 250 cycles, but the blank Na||S cell failed almost immediately. Zhu et al. prepared a mixed SEI layer with NaCl, PhS2Na2, and Na2S via the in situ reaction of p-dichlorobenzene and S8 on the surface of Na anode.161 The PhS2Na2-rich SEI layer showed a reduced binding energy (−2.3 eV) for Na+, promoting uniform Na deposition behavior in the carbonate electrolyte. According to DFT calculations, the binding energy between the components of the SEI membrane generated in carbonate solvent (−3.515 eV, −2.497 eV, and −2.49 eV, respectively) had a small force, which may speed up Na+ transport in PhS2Na2. A full battery assembled with a Prussia Blue (PB) cathode and bare Na anode displayed a capacity retention rate of only 50% after 1000 cycles, while the PhS2Na2-protected Na anode with an initial capacity of 89.5 mA h g−1 still had a capacity retention rate of 93% after 1000 cycles.
Fig. 14 (a and b) Young's modulus test of PNF and PVDF. (c and d) The cycle performance of the symmetrical battery with different treated electrodes at a current density of 1 mA cm−2 and area capacity of 2 mA h cm−2 and 3.2 mA h cm−2, respectively. (e) Comparison of the cycle performance of Cu||FeS2, PVDF@Cu||FeS2, and PNF@Cu||FeS2 at 0.2 A g−1.162 (f) The cycling performance of Na plating/stripping on Cu and PVDF@Cu current collectors at a cycling capacity of 1 mA h cm−2 at 2 mA cm−2; (g) voltage profiles of symmetrical battery with bare Cu and PB@Cu collectors at 1 mA cm−2 with a fixed cycling capacity of 2 mA h cm−2.163 (Reprinted with permission from ref. 162 and 163. Copyright 2019 Elsevier B.V. and Copyright 2021 Elsevier B.V.). |
Na@SEI | Electrolyte | Current density (mA cm−2) | Energy density (mA h cm−2) | Symmetrical battery (cycle life@overpotential (mV)) | Ref. |
---|---|---|---|---|---|
Na@NSCNTs | 0.5 M NaTFSI + diglyme | 1 | 1 | 500 h@70 | 31 |
Na@Al2O3(ALD) | NaPF6 + diglyme | 3 | 1 | 100 h@17 | 151 |
Na@Al2O3(PEALD) | 1 M NaClO4 + EC + DEC | 0.25 | — | 450 h@100 | 152 |
Na@NaxSny + NaCl | NaTFSI + diglyme | 2 | 1 | 325 h@— | 153 |
Na@Na-Bi | Carbonate-based electrolyte | 1 | 1 | 500 h@20 | 154 |
Na@Na3Bi + Na3ClO | Carbonate-based electrolyte | 1 | 1 | 700 h@30 | 155 |
Na@NaOH + NaNH3 | 1 M NaTFSI + diglyme | 1 | 1 | 1000 h@— | 158 |
Na@NaF | 1 M NaTFSI + diglyme | 1 | 1 | 1000 h@8 | 160 |
5 | 5 | 600 h@— | |||
Na@Na2S + NaCl + R-S-Na + PhS2Na2 | EC + PC | 1 | 1 | 800 h@15 | 161 |
Na@PVDF | 1 M NaPF6 + diglyme | 1 | 1 | 1200 h@35 | 32 |
Na@PVDF + NaF | 1 M NaPF6 + diglyme | 1 | 2 | 900 h | 162 |
1 | 1 | 2100 h | |||
Na@Na3Bi + PVDF | 1 M NaPF6 + diglyme | 1 | 1 | 2700 h@35 | 163 |
1 | 2 | 2500 h@— |
The Na/solid-state electrolyte interface can also be improved by the modification of Na anodes. Luo et al. added amorphous SiO2 into Na metal to construct a Na–SiO2 composite (Fig. 15a), which can reduce the surface tension and enhance the contact area between Na+ superionic conductor (NASICON)-based SSEs and Na anode.164 It was discovered that the interface resistance of Na||NASICON was 1658 Ω cm−2, much more than that of the Na-SiO2||NASICON (just 101 Ω cm−2) in Fig. 15b. At a current density of 0.1 mA cm−2, the original Na anode stopped working after 9 h, but the Na–SiO2 composited anode showed a longer stability, indicating that reducing the interface resistance and enhancing the interface contact between the electrode and the electrolyte are crucial to the operation of SSEs (Fig. 15c). Cheng et al. annealed Na–K alloy to attach the carbon fiber at 750 °C to obtain the carbon-fiber-supported liquid Na–K alloy.165 The carbon fiber would generate flaws on the structure after high temperature treatment, increasing its affinity with Na–K alloy (record as ZNSE). The alloy electrode and NASICON were joined at a pressure of 50 MPa, and the Na–K alloy was entirely bonded to the SSEs with no visible cracks or dendrites. At 1 mA cm−2, the polarization voltage of Na–K||ZNSE||Na–K was much lower (which remained at 29 mV during cycles over 900 h) than that of Na||ZNSE||Na (which increased from 81 mV to 128 mV). The Na–K alloy anode and NVP electrode were combined to make a whole battery that could cycle for more than 100 times under ideal conditions. Therefore, it is promising to improve the ionic conductivity and interface compatibility between SSEs and Na anode by modifying the surface of electrodes, thus promoting the practical applications of SSEs in SMBs.
Fig. 15 (a) Contact schematic between the SSEs and Na or Na–SiO2 electrode. (b) The impedance measurement of Na and Na–SiO2-based battery. (c) The cycle performance of symmetrical cells using various electrodes at 0.1 mA cm−2.164 (Reprinted with permission from ref. 164. Copyright 2019 American Chemical Society). |
Fig. 16 (a) Schematic representation of how the NZSP surface structure changed in the presence of water (H2O + CO2) and elevated temperatures. (b) Evaluation of EIS schematics for symmetrical Na/NZSP/Na and Na/NZSP (heat treatment)/Na batteries. (c) Discharge/charge cycle analysis of symmetric Na/NZSP/Na and Na/NZSP (heat treatment)/Na batteries at 0.1 mA cm−2.166 (d) Diagrammatic representation of the anticipated reaction pathway for photoinduced crosslinking during XPE preparation. (e) The stability of the XPE interface in a symmetric Na/XPE/Na arrangement at room temperature. As seen in the illustration, the resistance number is taken from a single Nyquist diagram that the EIS measured.168 (Reprinted with permission from ref. 166 and 168. Copyright 2020 American Chemical Society and Copyright 2017 Elsevier B.V.). |
Unlike oxide electrolytes, sulfide electrolytes will produce many side reactions with Na anodes on the interface; thus, the SEI layer prepared in sulfide SSEs needs to have certain chemical inertness to prevent the side reaction with sulfide from destroying its own structure. One of the solutions proposed by researchers is to add an interlayer between the SSE and the Na anode to improve the wettability and Na+ conductivity of the interface. For example, molecular deposition technique (MLD) is used to produce an alucone film layer on the surface of Na anode.167 A layer of alucone film was produced by 150 deposits on the surface of Na anode. The treated electrolyte was used to assemble a symmetrical battery to test its polarization. The Na@mld150C-Na3SbS4-Na@mld150C symmetrical battery could cycle for 475 h with a final polarization voltage of 450 mV. The polarization voltage of Na–Na3SbS4–Na battery increased to the value of 1.2 V from the initial 56 mV, with a cycle life of only 260 h. This method effectively separated the Na anode from the sulfide SSEs (Na3SbS4 and Na3PS4), preventing the breakdown of sulfide-based electrolytes and the formation of Na dendrites. This completely demonstrated that the alucone film had significant advantages in suppressing Na dendrite and stabilizing battery cycle performance.
Gerbaldi et al. developed a crosslinked polymer electrolyte (XPE) based on poly(ethylene oxide) (PEO) as the substrate, which was prepared by a solvent-free photopolymerization technique (Fig. 16d).168 This XPE with NaClO4 showed a high conductivity (>1 × 10−3 S cm−1) at room temperature, and a passivation layer between the XPE and the electrode increasingly formed during the battery cycle. In the succeeding operation, this passivation layer will cause a modest rise in impedance and circulate steadily for 200 h (Fig. 16e). In addition, the therm gravimetric analysis test indicates that at high temperatures, the PC in the XPE volatilizes substantially less than that in the noncrosslinked electrolyte. Li et al. used IL electrolyte PYR14TFSI blended with NaTFSI (PYR/NaTFSI) as an effective interlayer to stabilize the interface of Na metal and sulfide SSE. This PYR/NaTFSI layer favored Na+ transfer and solved the high interface resistance generated by the solid–solid contact.169 When the current density was 0.1 mA cm−2, the Na3SbS4-PYR/NaTFSI electrolyte in the symmetric cells can be used for stable cycling for 300 h, but the polarization voltage of the blank Na3SbS4 electrolyte suddenly dropped after 53 h, mainly owing to internal short circuit caused by Na dendrite formation.
Lee et al. developed a simple synthesis method by introducing nanocrevasses on the carbon fiber scaffold, which could make it easier for alkali metal penetrating into carbon cloth to improve the wettability of Na metal (Fig. 17a).181 To put together, the initial Na electrode and Na/C electrode were used. The polarization of the symmetrical cells with bare Na anode increased after 10 h at a current density of 3 mA cm−2, whereas the polarization of the Na/C anode in symmetrical cells did not substantially increase after 80 h. The Na/C anode displayed a capacity of 1106 mA h g−1 and a capacity retain of 94.8%. In contrast, only 80.9% of the potential capacity (942.8 mA h g−1) was released when bare Na foil was used. Fan et al. used commercial carbon felt as a feedstock by soaking it into the molten Na liquid to obtain Na/C composites, producing an electrode with great mechanical strength, electrochemical properties, good thermal stability, and substantial surface area.33 As the carbon scaffold served as the site of Na deposition during the electrochemical cycle, Na plating and stripping capacities were kept constant at 2 mA h cm−2 in symmetrical cells. Symmetrical cells with bare Na demonstrated a large Na stripping/plating overpotential (>40 mV) and considerable irregular fluctuations throughout the cycle at 1 mA cm−2. The Na/C composites not only displayed a significantly reduced overpotential (20 mV) but also attained stable cycle performance after 120 cycles, which even showed a small overpotential (50 mV) at 3 mA cm−2. Besides, Yan et al. used 3D printing technology to prepare a reduced graphene oxide/carbon nanotube (rGO/CNT) microcrystalline aerogel with micropores and adjustable layered microcrystalline for Na metal host.182 This rGO/CNT aerogel decreased the growth of Na dendrites and offered abundant active sites for Na deposition, facilitating the interface between the electrolyte and active ingredients, the penetration of the electrolyte into the nanostructures, and the ion transport rate. As a result, the CE of the main metal battery with rGO/CNT was maintained above 99%, while the initial CE of the blank Cu electrode in the symmetrical cell was only 48%, followed by a jump in the range of 83–100%.
Fig. 17 (a) Schematic diagram of the manufacturing process of Li and Na metal/carbon composites.181 (b) Schematic illustration of the preparation process of 3D Au/CNT-GF nanostructures. (c) The current density distribution on Cu, 3D CNT-GF, and 3D Au/CNT-GF electrodes simulated by COMSOL Multiphysics. (d) Polarization voltage curves of CuNW-Cu and bare Cu assembled asymmetrical cells.63 (Reprinted with permission from ref. 181 and 63. Copyright 2019 American Chemical Society and Copyright 2022 Elsevier Ltd). |
To promote the uniform distribution of Na+ on carbon-based materials, metal oxides can also be employed as active sites for Na deposition. Ye et al. used commercial carbon cloth (CC) modified by the concentrated nitric acid in RuCl3 solution to obtain the lithiophilic CC layer containing RuO2 nanoparticles (RuO2@CC).184 This Ru@CC layer showed good Na affinity during cycling, and the presence of RuO2 particles could further limit the current density and impede dendrite formation of Na anode. The Na-Ru@CC electrode in the symmetrical battery showed stable voltage profile for over 250 h at 1 mA cm−2 with 1 mA h cm−2 in 1 M NaClO4/EC with 5% FEC electrolyte, but the Na-Ru@CC electrode under the same conditions exhibited evident voltage fluctuations and sudden short-circuit at 58 h, originating from uneven plating/stripping on the surface of Na metal and severe dendritic growth.
Li+ is prone to gather near the negative surface because of its lower redox potential (−3.04 V vs. SHE) than Na (−2.71 V vs. SHE), which could form a strong electrostatic shielding effect to prevent dendritic growth during Na deposition on the Li–Na alloy anode (Fig. 18a).187 The liquid Li–Na alloy instantly covered the carbon cloth (the Li/Na molar ratio was 2:1), creating a LiC6-based interfacial coating. The electrochemical behavior of LiNa-CC surfaces was distinct from that of Li and Na surfaces due to the eutectic melting of Li and Na alloy crystals, respectively (Fig. 18b and c). The symmetrical battery with a LiNa-CC-based anode demonstrated improved CE and superior electrochemical performance (maintained for 700 h cycles and a polarization voltage of 56 mV at 2 mA cm−2) as compared to the blank Na anode, which had a lifetime of less than 150 turns (Fig. 18d and e). Liu et al. mechanically mixed Na75K alloy nanorods with PTFE particles, which underwent a crosslinking reaction during the mixing process, creating an Na metal anode with a 3D noncollapsing frame.188 The nanostructured Na75K caused the defluorination of PTFE, leading to the formation of the crosslinked artificial SEI with NaF and KF. The PTFE@NaK//PTFE@NaK symmetric cell showed 60 mV micropolarization after 707 cycles, while the blank Na||Na symmetric cell exhibited an increased polarization voltage after the initial cycle and failed within 100 cycles.
Fig. 18 (a) Schematic diagram of LiNa-CC anode preparation. (b) Schematic diagram of the symmetrical battery with various metal anodes. (c) The schematic illustration of liquid LiNa shell with enhanced electrochemical performance.187 (Reprinted with permission from ref. 187. Copyright 2021 The Royal Society of Chemistry). |
The formation of Na dendrites can be reduced and the long-term cycling of SMBs can be promoted by the combination of alloy formation between alkali metals, in view of the electrostatic shielding effect and the reduction of current density. Besides the alloying between alkali metals, Na alloys with excellent properties can also be obtained using nonalkali metals. Huang et al. added a permeable ion/electron conductive framework via the reaction between molten Na and SnO2, forming a percolated Na–Sn alloy/Na2O framework (NSCA-31) on the Na metal (Fig. 19a and b).189 The exposed “sodiophilic molecule” framework and its quick electron transport capabilities caused a uniform distribution of local current during the subsequent electroplating, which made it easier for Na+ to deposit into the interior space while dampening the volume variations. The plating/stripping dynamics and stable interface of the NSCA-31 electrode (Na/SnO2 weight ratio of 3:1) resulted in the resistance of the electrode remaining low and steady even after 50 cycles. It was clearly found that the treated electrode assembly had a smaller polarization and a longer cycle life than that of bare-Na assembled battery (Fig. 19c and d).
Fig. 19 Rapid and uniform Na+ stripping and dendrite-free Na+ plating on Na metals with ion/electron conductive frames. (b) Slow/uneven Na+ stripping and dendritic Na+ plating on bare Na metal. Cycling performance of symmetrical cells with bare Na and NSCA-31 electrodes with a capacity of 1.0 mA h cm−2 and the current densities of (c) 1.0 mA cm−2 and (d) 2.0 mA cm−2.189 (Reprinted with permission from ref. 189. Copyright 2019 Elsevier Ltd.). |
Moreover, Al is chosen as the candidate collector due to its high electrical conductivity and inert properties (Na is not alloyed with Al). Wang et al.191 reported a porous Al collector as an electroplating substrate to inhibit Na dendrite growth. By reducing the distribution of Na flux and increasing the surface area available for Na nucleation, the interconnected porous structure may achieve uniform deposition. As a result, the Na anode with porous Al collector showed a steady plating/stripping performance for 1000 h, with an average CE of 99.9%. By contrast, the battery with the flat Al foil had a short circuit and a growing voltage lag. Besides, because Ni metal and Na metal have good bonding ability, it is also a good choice to use Ni metal to prepare a 3D frame for Na anode. Lu et al. used the mechanical rolling strategy to immerse Na metal in 3D Ni foam to obtain a 3D Ni/Na anode, which was favorable for decreasing the growth of Na dendrites caused by an excessive current density.192 The 3D Ni/Na anode exhibited a polarization voltage of 13 mV and a stable cycle of 600 h at a current density of 1 mA cm−2. In addition, the 3D Ni/Na anode could show a low polarization voltage of 38 mV at 3 mA cm−2, and 3D Ni/Na||NVP full cells could stably cycled for more than 200 cycles with an initial capacity of 94 mA h g−1.
Similarly, employing a metal oxide frame could also lower the surface current density to induce the uniform Na deposition and prevent the growth of Na dendrites. Li et al. reported the self-supported S–TiO2 nanotube arrays via the electrochemical anodization and subsequent sulfidation process (Fig. 20c).193 The S–TiO2 tubular array has outstanding impedance characteristics when used as a binder-free electrode in an Na cell (Fig. 19d). The Na storage efficiency of S–TiO2 can be attributed to its distinct nanotube array properties, significant doping effect on electronic properties, and enhanced dynamic stability. The S–TiO2 nanotube with an outer diameter of 80 nm contributed to the even stripping/plating behavior of Na metal, enabling a long-term cycle stability (with a capacity retention of 91% after 4400 cycles) for the SMBs. The S–TiO2 nanoarray electrode underwent a long-term test (Fig. 20e).
Fig. 20 (a) Schematic diagrams of Na nucleation and deposition on Cu foam (left) and CF@ZnO (right). (b) Rate performance of symmetric cells with different Na anodes.34 (c) Electrochemical impedance spectra of S-TiO2 and TiO2 nanotube array anodes at different cycle stages. (d) The EIS spectra of the S-TiO2 nanotube array assembled anode after 2500 cycles at 5 C. (e) Comparison of cycling performances of the S-TiO2 nanoarray and TiO2 electrodes at 5 C.193 (Reprinted with permission from ref. 34 and 193. Copyright 2016 WILEY-VCH and Copyright 2020 Elsevier B. V.). |
Yang et al. reported a 3D porous Cu foam containing ZnO nanorod arrays (CF@ZnO).34 As shown in Fig. 20a, Na is typically sparsely distributed along the Cu ligaments’ bones, particularly on their curved ridges. The irregular island architecture and high electric field distribution of the Na turning on bare Cu foam led to the gradual transformation of Na particles into Na dendrites and moss Na. The sodiophilic ZnO nanorod arrays on porous Cu form provided abundant nucleation sites for Na metal, showing a low polarization voltage of 82 mV during deposition and stripping, whereas the untreated Na anode exhibited a higher polarization voltage of 207 mV than that of CF@ZnO. The full cell with CF@ZnO/Na anode delivered an outstanding rate capability, which showed that the CF@ZnO/Na anode could induce Na+ deposition well.
Fig. 21 (a) Schematic diagram of the synthesis route of h-Ti3C2/CNTs.62 (b) Schematic diagram of the manufacturing procedures for the Ti3C2Tx-CC frame and Na-Ti3C2Tx-CC metal anode.195 (c) Schematic illustration of the synthesis process of the Ti3C2Tx@Sb sample.196 (d) XRD spectra of CT-Ti3C2 and CT-Sn(II)Ti3C2 before and after Sn2+ addition. (e) Schematic diagram of the preparation of CT-Sn(II)@Ti3C2 by CTAB pretreatment and Sn2+ column brace method.198 (Reprinted with permission from ref. 62, 195, 196 and 198. Copyright 2020 Wiley-VCH, Copyright 2020 American Chemical Society, Copyright 2019 IOP Publishing Ltd and Copyright 2018 WILEY-VCH). |
Moreover, the electrochemical properties of SMBs could be improved by the combination of MXene with carbon fiber. Cao et al. reported a MXene (Ti3C2Tx)-coated carbon cloth (Ti3C2Tx-CC) with high metal conductivity and sodiophilic surface.195 The flexible Na-Ti3C2Tx-CC electrodes could be manufactured by a thermal infusion treatment in Fig. 21b. As a result, this Na-Ti3C2Tx-CC foil could be folded and rolled indefinitely, making it advantageous for use in flexible or irregularly shaped batteries of flexible Na anodes. The Ti3C2Tx-CC induced uniform deposition and low interfacial resistance at 1.0 and 5.0 mA cm−2 current densities. The Ti3C2Tx-CC electrode showed a CE of 98.5% after 900 h, but the CC electrode showed a CE of 94% after cycling for 630 h.
Furthermore, MXene with columnar morphology can also be effective to lower the current density and prevent Na dendrite formation. Li et al. reported Sn2+-pillared Ti3C2 prefabricated by cetyltrimethylammonium bromide (CTAB) precolumnized with Ti3C2 (CT-Sn(II)@Ti3C2), which was used as a stable substrate for the dendrite-free Na metal anodes.198 After CTAB pretreatment, the layer spacing of Ti3C2 was increased to 2.2 nm, indicating that CTAB was successfully embedded between Ti3C2 layers. No additional peaks were observed in the XRD spectra of CT-Sn(II)@Ti3C2 after Sn2+ columned, suggesting the presence of amorphous intercalated Sn(II) nanocomplexes in the CT-Ti3C2 matrix (Fig. 21d). In addition, the CT-Sn(II)@Ti3C2 electrode could still obtain a high CE of 98.8% under a current density of 5 mA cm−2 and an areal capacity of 5 mA h cm−2. Impressively, CT-Sn(III)@Ti3C2 was able to provide a high CE of 98.5% after 100 cycles even at 10 mA cm−2, further demonstrating the positive impact of Sn2+ inserted between Ti3C2 layers on Na nucleation and deposition behavior.
In summary, the surface current density can be greatly decreased by incorporating a 3D conductive frame into the Na anodes, and the reduction in the current density successfully causes the uniform deposition of Na+. In addition, adding Na+ nucleation sites to the 3D frame or enhancing the sodiophilic surface of the 3D structure can enhance the battery performance as well.
Although Na metal anode has achieved amazing progress so far, there are still several obstacles in this field, including low CE, poor cycling stability, and limited energy density. This review puts forward the promising development prospects and research directions for Na anode protection, promoting the further applications of SMBs in practical large-scale energy storage devices. As an outlook, the structure/interface engineering on Na metal anode of SMBs can be further improved through the following aspects.
(1) Firstly, the formation mechanism and evolution process of Na dendrites should be deeply analyzed and monitored with high-resolution characterization techniques and advanced theoretical calculations. In particular, in situ characterization techniques (i.e., cryo-electron microscopy, high-resolution TEM, and TOF-SIMS) are promising tools to resolve the dendritic dilemma, speeding up the data analysis process of structural/interfacial changes of Na metal anodes. Moreover, the emerging machine learning methods should be used to illustrate the Na dendrite evolution mechanisms in the near future.
(2) Secondly, the types of anodes for SMBs should be expanded and explored, and the related issues should be taken into consideration as well. For instance, the alloy-based electrode may be served as a potential replacement for a new generation of high-energy-density anodes. The composited anodes not only reduce the volume change but also increase the electronic conductivity. The utilization of polymer or hybrid organic and inorganic electrode materials offers significant potential for high safety batteries. Nevertheless, the mechanism of Na+ exfoliation and deposition in polymers and organic/inorganic interfaces requires additional investigation.
(3) Thirdly, the modification and optimization of the SEI layer should consider the components and structures of electrolytes, cathode, and anode, which are crucial parameters for constructing uniform SEI layers. Strategies introducing in situ SEI with electrolyte additives and constructing artificial SEI layer put promote the ion transfer rate and uniformity of ion distribution on the Na metal surface. Notably, atomic/molecular layer deposition methods could be effective ways for producing high-quality SEI films, which can provide a controlled surface protective layer.
(4) Fourthly, the accessible surface area of the 3D frame causes significant Na and electrolyte consumption during the repeated deposition and peeling of Na+ in the 3D frame. Na metal still contacts the electrolyte while being enclosed in the matrix. To effectively increase the stability of the Na anode, the protective coating must be formed on the surface of the three-dimensional structures. The combination of SEI modification and 3D frame may open up an avenue toward high-performance SMBs with enhanced stability.
(5) Fifthly, the construction of 3D conductive frameworks should take the gravity and volume energy densities into account, which may affect the actual energy density of the whole battery. The obtained electrode cannot make full use of the energy density of its volume and weight in practical applications at present, and the weight and volume of the designed framework and Na should be optimized. The quantitative relationship between the effects of Na dendrite suppression and the whole energy density of the SMBs need to be further studied.
Moreover, SSEs are proposed to address the electrolyte loss and safety issues of the traditional liquid electrolytes of SMBs. However, the initial solid–solid interface of SSEs causes an increase in the interface impedance, leading to declining CE and increasing polarization in the SSE-based SMBs. Notably, the interface compatibility of SSEs could be improved by the methods of SEI modification, structural transformations of SSEs, enhanced interfacial contact between SSEs and Na metal, etc. Therefore, the further exploration and structural design of SSEs with high ionic conductivity, well-interfaced compatibility, and high safety are effective means to promote the commercial application of SMBs.
As for the industrialization development trend of SMBs, cost is the main benefit of Na metal batteries over Li metal batteries. Firstly, compared to Li metal, Na metal is more widely distributed in the crust, which results in a lower price for Na metal anode. Secondly, Al foil can be used as the current collector in SMBs, which is less expensive than that of the Cu collector in Li metal batteries. Theoretically, SMBs have a higher energy density than sodium ion batteries, and they will have a wider range of potential applications in the future. Due to the rising popularity of electric vehicles, cost control will eventually become a significant barrier to the development of Li metal/ion batteries. Na metal/ion batteries can store more raw materials than Li metal/ion batteries, which provides a structural foundation for next-generation electric vehicles and future large-scale electrical storage. In addition, the Na metal/ion battery-powered devices could be used in conjunction with Li metal/ion battery-powered devices, which lowers the cost of enterprise transformation and encourages the manufacturing of SMBs.
In addition, in the field of low temperature batteries, it is also one of the advantages of the future industrialization of Na ion/metal batteries. Compared with Li ion/metal battery, the solvation energy and transport barrier of Na+ in the electrolyte are smaller than those of Li+, which could enable Na ion/metal batteries with higher coulombic efficiency at lower temperatures. In the field of low-temperature batteries, the minimum operating temperature of Li-ion batteries is −20 °C, while Na-ion batteries can work normally at a low temperature of −40 °C. This provides a solution for the application of low-temperature batteries in the future and the demand for batteries in some extreme environments, which will promote the industrialization of Na ion/metal batteries in the future.
Nevertheless, the electrolyte formula has a significant impact on electrode protection across the entire battery system. More comparative studies between ester and ether electrolytes should be added in the upcoming study of electrolyte formula. Comparative studies help to improve the mix of new electrodes and electrolyte solutions and to produce more stable and effective battery cycles, which could provide in-depth strategies to promote further commercial applications of SMBs.
Footnote |
† Celebrating the 25th anniversary of the Key Laboratory for Special Functional Materials of Ministry of Education at Henan University. |
This journal is © The Royal Society of Chemistry 2023 |