A Review of Post-Lithium-Ion Batteries

A Review of Post-Lithium-Ion Batteries By Laurence Middlemiss & Alex Holland, PGR students, EPSRC CDT in Energy Storage & Its Applications Introduction Rechargeable batteries are characterised by their energy density (Wh kg -1 ), specific energy (Wh L -1 ), specific power (W kg -1 ), cycle life, efficiency and cost. This in turn is determined by the performance of the components that make up the battery cell: negative and positive electrodes as well as separator and electrolyte. Importantly, the capacity and voltage of the electrodes determine the energy density and specific energy of the cell, as seen in equations (1) and (2). C cell = C ac c C a +C c (1) Energy density = C cell V cell (2) Additionally, these characteristics change depending on how the cell is used, e.g. rate and depth of cycling. Depending on the application, focus can be given to the parameters most important for that given application. The lithium-ion (Li-ion) battery is ubiquitously used in consumer electronics due its high energy density. It functions through the intercalation/de-intercalation of Li + ions into and out of a positive and negative electrode, with different Li-ion chemistries generally categorised by the positive electrode employed, resulting in a range of performance characteristics. The majority of modern cells consist of a graphite negative electrode and a metal oxide positive electrode such as lithium iron phosphate (LiFePO 4) or lithium cobalt oxide (LiCoO 2). A polymer soaked in a liquid lithium electrolyte is usually used to separate positive and negative electrode and allow the transport of Li + between them. The permeable separator allows ions to pass through, but is electrically insulating, blocking electrons, and forcing them through the external circuit. The half- and full-cell reactions of a LiCoO 2 cell, the chemistry most commonly used in current consumer electronics, are given by reactions (3), (4) and (5). Reactions are driven to the right during charge. C 6 + Li + + e - LiC 6 ca. 0.05 V vs Li/Li + (3) LiCoO 2 CoO 2 + Li + + e - ca. 3.8 V vs Li/Li + (4) C 6 + LiCoO 2 LiC 6 + CoO 2 ca. 3.85 V (5) A comparison of the performance metrics of different lithium-ion battery electrodes is given in Table 1. The performance requirements of electric vehicles (EV), and decarbonisation of power grids, mainly drives the development of Li-ion and post Li-ion battery technologies. For EVs, high specific energy is the primary requirement in order to increase the range of the vehicle while cycle life must remain high enough to allow use throughout the projected lifetime of the vehicle (ca. 10 years). Cost reductions may also be required for widespread adoption as the battery pack can account for up to 40% of the EV cost [1]. The requirements for grid use, however, can vary drastically depending on application. Parameters such as lifetime cost, specific power and cycle life can take precedence. 1

Table 1. Performance metrics of different lithium-ion battery electrode materials. Electrode material Capacity (practical) Potential vs Li/Li + Cell cycle life (vs graphite) C-rate Graphite 372 mah g -1 (300-350 mah g -1 ) ca. 0.05 V N/A N/A LTO 150 mah g -1 1.55 V 10,000+ 10C LiCoO2 170 mah g -1 3.8 V 500-1,000 1C LiFePO4 130-140 mah g -1 3.4 V 2,000 5C Lithium-Ion Generally, Li-ion R&D focusses on improvements to energy density and specific energy, which would benefit large range EVs and long life electronic devices. These huge markets ensure that new high energy chemistries will be mass produced, thus bringing down costs and potentially allowing their use in other applications. LTO electrodes. Lithium titanate (LTO) cells incorporate Li 4Ti 5O 12 in place of graphite at the anode. Compared to graphite, LTO has a low theoretical capacity of 150 mah g -1 and a more positive Li intercalation potential of approximately 1.55 V, leading to relatively low energy densities and specific energies [2-7]. The half-cell reaction is given by the reaction below. Li 4Ti 5O 12 + 3Li + + 3e - Li 7Ti 5O 12 ca. 1.55 V vs Li/Li + (6) Conversely, the more positive lithiation potential ensures dendrite and solid electrolyte interphase (SEI) formation is avoided and improves the safety of high rate cycling. The solid electrolyte interphase is formed on the surface of the anode during the first charge, when an increase in voltage causes species in the electrolyte to be reduced. While a stable SEI layer is necessary for the successful functioning of a cell, its formation must be controlled, as over time, too thick an SEI can lead to performance degradation and capacity loss. As such, nano-lto can be used to further improve rate capability since the increased surface does not cause excessive SEI growth as is common for other nano Li-ion electrodes [3-5, 8, 9]. Furthermore, LTO cells can exhibit exceptional cycle lives a 67.5 Wh kg -1 LTO/LFePO 4 18650 cell was capable of 30,000 cycles at 15C charge 5C discharge rate, with 91% capacity retention [10], 2

Figure 1. This stems from the very small 0.2%, volume change that occurs during cycling, significantly reducing the stress fractures that plague graphite anodes. Hence, LTO is often referred to as a zero strain electrode [3-6]. Figure 1. (Left) Typical voltage profile of an LTO anode vs. Li/Li +. (Right) Capacity retention of a 67.5 Wh kg -1 LTO/LFePO 4 18650 cell [10, 11]. While initial cost of LTO cells are higher than other Li-ion chemistries, the cyclability of the chemistry is likely to result in a superior $/kwh over its lifetime, while environmental impacts are also favourable for LTO [12, 13]. Despite low energy density, which leads to increased initial cost/kwh, the improved thermal stability [14], rate capability and cycle life of LTO make these type of cells well suited to a number of applications, such as various grid applications or the electrification of public transport and taxis. As such, LTO has been successfully commercialised by a number of companies, e.g. Toshiba. Silicon anodes. Silicon anodes have been widely researched as an alternative to graphite as a negative electrode. The electrode functions through the formation of the alloy Li 15Si 4 at room temperature (higher capacities can be obtained at higher temperatures) resulting in high theoretical capacities of 2,400 mah cm -3 and 3,590 mah g -1, which is approximately an order of magnitude greater than the theoretical specific capacity of graphite [15-18]. The alloying process follows reaction (7) with a nominal Li intercalation voltage of approximately 0.4 V, theoretically providing an energy density increase of 30% over a LiCoO 2 cell. 15Li + + 4Si + 15e - Li 15Si 4 ca. 0.4 V vs Li/Li + (7) However, improvements on this scale are unlikely to be realised. This is because of a near 300% volume expansion during lithiation (charge), Figure 2. Repeated volume changes during cycling results in the silicon particles fracturing, which leads to a loss of electrical contact between electrode constituents and SEI instability, all of which severely limit cycle life and lower coulombic efficiency. 3

Figure 2. Si electrode failure mechanisms: (a) material pulverization. (b) Morphology and volume change of the entire Si electrode. (c) Continuous SEI growth. [19] A number of strategies have been employed to improve the stability of Si anodes, including the use of nanostructures, void spaces, conductive coatings, structural networks and intelligent binder choice [15, 20, 21]. Nanoscale Si is almost exclusively employed in order to minimise volume change per particle, while the introduction of void spaces within particles has been shown to be particularly effective in reducing mechanical stress by allowing space for volume changes [15]. Lifetimes over 1,000 cycles have now been achieved by Si anodes employing void structures. A capacity of 600 mah g -1 was possible for 6,000 cycles at 12C using a double-walled silicon nanotube anode [22]. At a 1C cycle rate, capacity was approximately 1,250 ma h g -1. Coulombic efficiency was also impressive at nearly 100%. A pomegranate inspired carbon coated Si nanoparticle cluster produced a capacity over 1,000 mah g -1 for 1,000 cycles (C/2 rate) and high coulombic efficiency (99.94%) [23]. A volumetric capacity of 1,270 mah/cm 3 is also promising, although it was unclear whether this was calculated in a charged or discharged state, while the majority of publications describe results in mah g -1, such that the effect of void introduction and nanoscaling on volumetric capacity is generally unclear, despite being of paramount importance. Additionally, many nanostructured electrodes employ low mass loadings (mg cm -2 ) and this, therefore, presents significant difficulty in assuming performance to be scalable on a cell level. Clearly, improvements to Si electrode lifetime have been possible through the use of nanostructures incorporating void spaces, but it should be noted that these novel nanostructures may prove difficult to synthesise at large scale and low cost. The use of Si in commercial cells has already begun with the incorporation of small quantities of SiO x powder, due to its relative ease of mass production. Typically, <5 wt% is added to the graphite anode [20]. The use of the most common PVDF binder will likely change with increasing SiO x quantities as performance is inferior compared to alternative binders with greater Young s moduli, such as polyacryllic acid (PAA), alginate, and carboxymethyl cellulose (CMC) [21, 24]. The carboxyl groups present in PAA and CMC allow contact to be maintained between anode particles and binder throughout cycling. These binders are also water soluble, thus negating the need for toxic and more 4

expensive NMP solvent. While it has a lower initial capacity than Si, SiO x results in the formation of lithium oxides during initial cycling, which can act as a buffer to volume expansion in subsequent cycles [24, 25]. However, the formation of these oxides naturally results in a low initial coulombic efficiency, due to the consumption of Li to form the oxide, such that excessive cathode loading is required, reducing cell energy density. Pre-lithiation of both Si and SiO x will be important for raising initial coulombic efficiency to minimise excessive cathode loading required in full cell arrangements [15, 24] (papers generally describe studies with a Li counter electrode such that there is effectively an infinite supply of Li + ). High voltage and high capacity cathodes. In order to increase energy density, a number of candidate materials have been investigated for their potential to operate at voltages up to 5 V vs Li/Li +. Polyanionic phosphates, sulphates and manganese oxides have all been investigated [26, 27]. An ordered spinel, LiNi 0.5Mn 1.5O 4 (LNMO), has been widely researched due to its flat nominal voltage profile at around 4.7 V and a practical capacity of 120-130 ma h g -1, which would amount to a specific energy increase of approximately 17% over graphite/licoo 2, Figure 3 [28]. For reference, the commercial LiCoO 2 cathode operates at a nominal discharge voltage of around 3.8 V, one of the highest voltages of commercial chemistries, with a reasonable practical capacity of approximately 140 ma h g -1. LNMO exhibits good rate capability [29], good safety, and uses relatively cheap and environmentally benign manganese, but currently suffers from severe capacity fade at temperatures around 60 C. [27]. The high rate capability of LNMO may be very important where fast charging is envisaged as a requirement for EVs. Figure 3. Charge/discharge profiles between 3.5 5.0 V for LiNi 0.5Mn 1.5O 4 hollow microspheres [30]. The primary limitation of high voltage cathode use relates to the oxidation and decomposition of common organic electrolytes. Secondly, the dissolution of transition metal cations from cathodes is exacerbated at high voltages. These issues can severely limit cycle life, cathode performance and safety. As such, the development of high voltage cathodes goes hand-in-hand with the development of a stable cathode-electrolyte interface [26-28]. The addition of organic and inorganic additives to conventional electrolytes has been employed to allow the formation of a stable SEI layer on the 5

cathode. This improves cycle life through shielding the electrolyte from transition metals in the cathode, which catalyse the decomposition of the electrolyte. Surface coatings, such as Al 2O 3, have also been widely investigated for LNMO cathodes due to their ability to suppress electrolyte oxidation and metal cation dissolution [27]. The development of solid electrolytes with high Li conductivity would be a highly desirable development. Further improvements are necessary for the stable cycling of high voltage cathodes in full cell configurations. As such, short term improvements to energy density may come from increasing the nickel content of the LiNi 1-x-yCo xmn yo 2 (NCM) cathode, LiNi 0.33Co 0.33Mn 0.33O 2, as well as to LiNi 0.5Co 0.2Mn 0.3O 2, which is already in commercial use [31]. At 0.1C, initial capacity can increase from approximately 160 ma h g -1 for LiNi 0.33Co 0.33Mn 0.33O 2 to 205 ma h g -1 for LiNi 0.85Co 0.075Mn 0.075O 2. However, the high reactivity of Ni 4+ leads to parasitic reactions and low thermal stability. The similarity in size between Ni 2+ and Li + results in blocking of Li + diffusion pathways by Ni 2+ while volume changes also increase at higher Ni content, all of which affect electrochemical performance. By the 100 th cycle, at 0.5C, a capacity of 133 mah g -1 from LiNi0.85Co 0.075Mn 0.075O 2 is inferior to the 140 ma h g -1 measured from LiNi 0.33Co 0.33Mn 0.33O 2, which can retain >80% initial capacity for over 3000 cycles [32]. At high Ni contents (x+y<0.4), data showing greater than 500 cycles is rare [33-37]. Rate capability and thermal stability also suffer at higher Ni percentages [33]. For example, the capacity of LiNi0.8oCo 0.10Mn 0.10O 2 (NCM811) can drop approximately 20% from 0.2C to 2C with LiNi0.4oCo 0.30Mn 0.30O 2 NCM433 only decreasing by approximately 9% [38]. Surface coatings and dopants are being used to limit parasitic side reactions and to mitigate migration of Ni 2+. Surface coatings were shown to improve the performance of NCM622 and NCM811 [39, 40]. NCM622 was treated with a Li-Co acetate in ethanol solution and NCM subject to an initial overcharging regime. Both treatments created layers where cation mixing took place, inhibiting further cation migration in the bulk cathode. However, only 150 cycles for NCM622 and 500 cycles for NCM811 were possible before an approximate 20% drop in capacity. The surface coating for NCM811 also lowered initial capacity to a comparatively low value of 155 mah g -1 at 100 ma g -1. An aluminium doped LiNi 0.84Co 0.06Mn 0.09Al 0.01O 4 was able to retain 94% of an initial 210 mah g -1 capacity at 100 ma g -1 [41]. Furthermore, promising results have been reported for Ni-rich NCM utilising a concentration gradient of cations within cathode particles, i.e. increasing the Mn content near the surface of particles to enhance stability with higher Ni content within the article to provide high capacity. Incorporating this concentration gradient for LiNi 0.65Co 0.13Mn 0.22O 2 produced capacities of 200 mah g -1 and 170 mah g -1 at 0.1C and 2C respectively, while full cell cycling at 1C provided 88% capacity retention over 1500 cycles [42]. Increasing the nickel content beyond 0.6 in NCM cathodes provides a realistic route to increased energy density, however, combining this with sufficient cycle life, rate capability and thermal stability still provides a challenge. Beyond Ni rich cathodes, lithium rich oxides could provide additional capacity improvements. Li 2MnO 3 has a theoretical capacity of 458 mah g -1 with nominal voltage around 3.7 V, while Li-rich LNMO and Li-rich NMC have been widely investigated, e.g. Li 1.2Ni 0.18Mn 0.59Co 0.03O 2, Li(Li 0.17Ni 0.25Mn 0.58)O2 [43-48]. Specific capacities of around 250 mah g -1 are commonly reported for these type of materials. However, as with Ni-rich layered compounds, the extraction of Li can result in the transformation of the layered structure into spinel phase, in addition to cation migration and dissolution [20, 48, 49]. Therefore, cycle life is often severely limited with the Li-rich cathode characteristic of voltage fade during repeated cycling (cation migration and trapping are proposed as the underlying cause of voltage decay). Surface coatings and modifications, with metal-phosphates and aluminium oxides, are current methods to improve stability by reducing interfacial reactions between the cathode and the 6

electrolyte [43-46]. Despite improvements, cycle life and rate capability are still considerably short of the performance required for commercial use. Lithium-Sulfur The employment of the sulfur element in batteries is desirable, due to its high theoretical gravimetric capacity of 1,675 mah g -1, as well as low cost and high availability [50]. However, the low volumetric capacity of sulfur may limit its potential use in EVs and personal electronics, instead making the chemistry better suited to certain grid or even aviation applications. Given 40% of cell cost can arise from cathode materials in Li-ion chemistries, replacement with sulfur, which is both lower cost and has a higher specific capacity, could feasibly allow for cost reductions [51]. The lithium-sulfur (Li-S) battery is now at the early stage of commercialisation. The sulfur cathode functions through a complex multi-step conversion reaction, while the anode is generally Li metal which undergoes plating and dissolution. The reversible sulfur lithiation is given by reaction (8) S 8 + 16Li + + 16e - 8Li 2S ca. 2.2 V vs Li + /Li (8) Historically, research has focussed on the sulfur cathode, due to its low electrical conductivity, 80% volume expansion upon lithiation, and complex reaction mechanisms [50, 52]. During lithiation, an intermediate and highly soluble polysulfide can migrate between electrodes, resulting in a loss of active material at the anode and passivation of the cathode. Inhibiting this process has been the subject of extensive research [20]. Using porous carbon materials as a host for sulfur can both increase conductivity and contain polysulfide species within pores, which can also help accommodate the 80% volume expansion [52-55]. Pore sizes smaller than 0.5 nm can completely eliminate the migration of polysulfides, but lead to low sulfur content an important factor when considering cell level design. Micro- (0.5-2 nm) and meso- (> 2 nm) pores are beneficial for allowing high sulfur content, but less effectively trap polysulfides, and so suffer from poor cycle stability http://onlinelibrary [50, 56]. While significant improvements have been made with regards to polysulfide migration and the sulfur cathode, extended cycling of a Li-S cell is still likely to be limited by the Li anode. Recently, focus has begun to shift toward the stability of the anode, although the majority of publications still concern the sulfur cathode. One issue regarding anode stability is its corrosion by polysulfides, which can still migrate to the anode despite efforts to contain them. Importantly, dendritic growth can occur at the Li anode, especially at high current densities. This results in the consumption of electrolyte and formation of powdery Li, leading to large polarisation and electrolyte consumption due to the increased surface area of powder and dendrites [52, 57]. Furthermore, the possibility of shortcircuiting stemming from this presents safety concerns. Electrolyte additives have been employed to increase anode stability through the formation of an SEI layer which helps suppress dendrite growth [56, 58]. LiNO 3, in ether based electrolytes, has been shown to be beneficial, but is insufficient to prevent parasitic reactions at high sulfur loadings, and can also decompose below 1.6 V at the cathode [52, 57, 59]. Interestingly, the addition of specific polysulfide species at specific concentrations can be beneficial for the formation of a stable SEI, although the changing concentration of species throughout cell operation makes implementing this difficult [57, 60, 61]. The use of functional separators, carbon interlayers, and surface coatings have all been investigated for their potential to improve anode performance through polysulfide trapping and dendrite suppression [62-66]. Cell capacities around 1,000 mah g -1 have been reported for several hundred cycles at >99% coulombic efficiency. Capacity fade at >1C rates can be severe, but given the high cell capacities, a 1C cycle rate still represents a relatively high specific power (ca. 2 W g -1 ). Maintaining the performance of research cells at larger scales, e.g. pouch cells, must now be demonstrated [67]. 7

Metal-Air Metal-air batteries promise a significant increase in specific energy due to the use of O 2 from the air as the active cathode material. While primary metal-air batteries exist, such as Zn-air or Al-air, electrically rechargeable metal-air batteries remain a significant challenge. Li-air is the most reversible chemistry (9) but still suffers from very poor cyclability, low efficiencies and poor rate capability [68]. Li + O 2 2Li 2O ca. 2.91 V vs. Li+/Li (9) As such, Li-air (and other secondary metal-air) batteries are far from commercialisation. Li-air cells can function in both aqueous and non-aqueous electrolytes with the majority of interest being in nonaqueous systems due to a higher theoretical energy density. The reversibility of Li-air cells is inhibited through a number of processes. In organic electrolytes, the inherent SEI formed on Li can suppress dendrite formation, but this decomposes due to repeated plating/dissolution, clogging cathode pores. Furthermore, the stability of Li in air due to moisture is low, which leads to large overpotentials, thus causing electrolyte decomposition and severely affecting cycle life. The choice/design of an electrolyte that is stable at both electrodes is seen as a primary research necessity, with room-temperature ionic liquids looking to be the most promising candidates so far. Deposits of electrically insulating Li 2O 2 must also be addressed as it will limit achievable capacity. Carbon black is commonly used as the air electrode. This limits the cycle life to 100 with 90% of initial capacity. The cathode solvent (e.g. NMP) and design have also been shown to be important for initial capacity and cycle life [69]. Energy efficiency can also be poor in Li-air batteries, due to overpotential between charge and discharge, e.g. 4.2 V charge to 2.5 V discharge at 50 ma g -1. Achieving the required power density and volumetric energy density in these kind of systems continues to be problematic. Most lab work is still focused on Li/O 2, rather than Li/air. Aqueous Li/air has many advantages but cell construction is likely to be complicated and challenging. There are serious questions over whether the lithium-air system will ever be realised as a practical reality. It is thought that the safety issues surrounding the organic electrolyte and lithium metal will be exacerbated if pressurised O 2 is required. For low cost, long range EVs, Li/air remains but is far from being a realistic option at present. Solid-State Batteries High voltage cathodes, Li-S and Li-air could all benefit hugely from the development of solid electrolytes which can provide improved electrochemical and thermal stability over liquid electrolyte counterparts. In addition to improved stability and safety on the cell level, this could also allow for simplifications of pack design and thermal management systems [70-72]. A lithium phosphorous oxy-nitride LiPON solid electrolyte was used in full cell configuration with Li anode and a thin film LiMn 1.5Ni 0.5O 2 cathode. At 5C, 10,000 cycles were performed, with a coulombic efficiency of approximately 99.98% and capacity retention of 90.6%. However, while the average cell discharge voltage at C/10 was around 4.6 V, at 5C the average discharge voltage dropped to around 4 V, with capacity also dropping from 120 to 90 mah g -1. In liquid electrolyte, capacity remained around 120 mah g -1 irrespective of cycle rate, while average discharge voltage dropped by round 0.15 V between 0.1C and 10C, although cycle life was significantly lower (depending on the excess volume of electrolyte used). Furthermore, the use of thin films is likely to significantly improve measured performance, with scalability a serious challenge. This helps to demonstrate the potential of solid electrolytes in improving stability, but also highlights the performance challenges that remain. 8

Namely, high rate performance can be poor due to low ionic conductivity through solid electrolytes at room temperature [73]. Other inorganic solid electrolytes with high conductivities exist but can instead suffer from low electrochemical potential stability, selective electrode compatibility, moisture sensitivity and/or poor mechanical properties. While polymer solid state electrolytes may be more amenable to large scale/size manufacture, many also suffer from similar issues, such as low ionic conductivity and difficulties in electrode compatibility [70-72]. Sodium-Ion Sodium-ion (Na-ion) batteries are viewed as a possible cheaper, safer, and more sustainable alternative to Li-ion technology. They are considered more sustainable as there is over 1,000 times more sodium than lithium in the Earth s crust, and additionally, sodium can be readily extracted from sea water [74]. This greater availability means that the metal is cheaper, and in addition, aluminium can be used as the current collector at the negative electrode, unlike Li-ion batteries which require more expensive copper [75]. This is because sodium does not react with Al to form an alloy. The technology can also be considered safer as sodium-ion batteries are less prone to thermal runaway they become self-heating at a higher temperature, and heat a lower rate [76]. Furthermore, Na-ion batteries can be stored at 0% of charge (fully discharged), making them much safer for transportation. Another huge advantage is that because the two monovalent-ion forming alkali metals have very similar chemistries, it is thought that, in the future, sodium-ion battery manufacturing can be easily based around what presently exists for Li-ion. However, there are a number of drawbacks to the development of sodium-ion batteries to be used as an alternative to Li-ion systems, some of which are laid out in Figure 4 [77]. Figure 4. A schematic illustration of the sodium-ion battery system and the challenges associated with its development [77]. 9

The greater size and weight of Na means that the theoretical capacity of the metal is lower [78], which makes it difficult to obtain materials with a rivalling energy density. Furthermore, the metal s lower reduction potential means that it is difficult to achieve as high cell voltages as seen with Li-ion batteries. There are also a number of challenges which need to be overcome with cathode and anode development, as well as a need to develop novel electrolytes. In a sodium-ion battery, the anode is usually some form of hard carbon, but not graphite, as the space between the layers of carbon sheets are too small to intercalate Na + ions. Similar with Li-ion batteries, the conductivity of solid sodium-ion electrolytes trails substantially behind those of liquid ones, so the most common electrolytes being researched are mixtures of either organic solvents or ionic liquids. A vast range of potential sodium-ion cathode materials have been explored to date, with the two most promising classes, which have received the most attention and development, being polyanionic materials and layered transition metal oxides [79]. The robust structure of polyanionic compounds means that they possess good structural stability, and have channels which allow facile Na + transport. Layered NaT MO 2 structures are relatively easy to synthesize, and have been found capable of high energy densities and voltages. The cathode Na(Ni 0.5Mn 0.5)O 2 is capable of a specific capacity of 125 mah g -1 (2.2-3.8 V), and even has a superior rate capability to its lithium counterpart [80]. However, there are questions surrounding its long-term cycling stability, which is a problem with these layered oxides in general, due to a large structural change on Na + (de)-intercalation. Improving the specific capacity and working voltage of sodium-ion batteries looks to be key to enhancing Na-ion development. One of the largest challenges is the volume change that occurs due to Na + ion (de)-insertion. So far, most research done on sodium-ion batteries to date has taken place on the cathode structure. In the future, it is envisaged that the major challenges that Na-ion technology will face will be largely the same as seen with Li-ion batteries now: cycle life and capacity fading. The importance of finding high-capacity anodes to compete with graphite in Li-ion systems is also crucial if sodium-ion batteries are to become a commercially relevant option. A recent cost analysis has shown that the lower energy density of sodium-ion batteries has not been compensated for so far [81]. It found that the energy normalised cost for lithium-ion and sodium-ion batteries are $0.11 and $0.14 per Wh respectively. Competing with high energy lithium-ion systems, such as lithium cobalt oxide, seems an impossible feat, especially for energy density. Therefore, rather than replace Li-ion technology, it is thought that sodium-ion batteries will become complimentary. Due to their larger size and reduced specific capacity, sodium-ion batteries are more likely to infiltrate stationary forms of energy storage (large-scale grid storage) long before mobile applications. Largescale stationary projects, where keeping cost down is imperative, is likely to be the main target market for Na-ion batteries going forward. It is hoped that if the above challenges are met, and desired requirements realised, then sodium-ion batteries may become commercialised in the near future (possibly the next five years) [82]. Multivalent-Ion (Magnesium-Ion/Aluminium-Ion) Secondary magnesium-ion batteries have the potential for high energy densities due to the ability of each Mg 2+ ion to store two electrons [83]. However, there is a major issue with finding a suitable cathode due to the large electrostatic interaction between the host anions and Mg 2+. This significantly limits the diffusion rate, and lowers electrochemical performance. 10

Al-ion batteries consisting of Al and pyrolytic graphite/carbon should be relatively cheap. Very high power has been demonstrated in such systems, with good cycle life (10,000), and these make use of RTILs, which are non-flammable [84]. However, low energy density/specific energy (due to difficulty of intercalating Al 3+ into a host cathode material) will limit applications and increase initial cost/kwh. Finding a suitable cathode is the primary objective of Al-ion and multivalent ion chemistries. RTILs are also difficult to handle (as with organic liquid electrolytes), and can be extremely corrosive, although this limits any passivation of aluminium. Conclusions In the short-term, improvements to Li-ion cathodes are likely to continue - providing incremental increases in energy density (and therefore a decrease in cost). However, beyond this there are much more exciting battery chemistries being looked at, which, if realised, could completely change the face of electrochemical energy storage. However, most of these, such as Li-air, are still a long way off commercial development. In the immediate future, sodium-ion and lithium-sulfur batteries pose the most realistic chance of coming into use, with companies in the UK already working on their development [85, 86]. 11

References 1. Bloomberg New Energy Finance, Lithium ion Battery Costs and Market. 2017. 2. Zhao, L., et al., Porous Li4Ti5O12 Coated with N-Doped Carbon from Ionic Liquids for Li-Ion Batteries. Advanced Materials, 2011. 23(11): p. 1385-1388. 3. Sandhya, C.P., B. John, and C. Gouri, Lithium titanate as anode material for lithium-ion cells: a review. Ionics, 2014. 20(5): p. 601-620. 4. Zhao, B., et al., A comprehensive review of Li4Ti5O12-based electrodes for lithium-ion batteries: The latest advancements and future perspectives. Materials Science and Engineering: R: Reports, 2015. 98: p. 1-71. 5. Sun, X., P.V. Radovanovic, and B. Cui, Advances in spinel Li4Ti5O12 anode materials for lithiumion batteries. New Journal of Chemistry, 2015. 39(1): p. 38-63. 6. Zhu, G.-N., Y.-G. Wang, and Y.-Y. Xia, Ti-based compounds as anode materials for Li-ion batteries. Energy & Environmental Science, 2012. 5(5): p. 6652-6667. 7. Nakahara, K., et al., Preparation of particulate Li4Ti5O12 having excellent characteristics as an electrode active material for power storage cells. Journal of Power Sources, 2003. 117(1): p. 131-136. 8. Yang, Z., et al., Nanostructures and lithium electrochemical reactivity of lithium titanites and titanium oxides: A review. Journal of Power Sources, 2009. 192(2): p. 588-598. 9. Amine, K., et al., Nanostructured Anode Material for High-Power Battery System in Electric Vehicles. Advanced Materials, 2010. 22(28): p. 3052-3057. 10. Zaghib, K., et al., Safe and fast-charging Li-ion battery with long shelf life for power applications. Journal of Power Sources, 2011. 196(8): p. 3949-3954. 11. Sandhya, C.P., B. John, and C. Gouri, Synthesis and electrochemical characterisation of electrospun lithium titanate ultrafine fibres. Journal of Materials Science, 2013. 48(17): p. 5827-5832. 12. Peters, J.F., et al., The environmental impact of Li-Ion batteries and the role of key parameters A review. Renewable and Sustainable Energy Reviews, 2017. 67: p. 491-506. 13. IRENA, Electricity Storage and Renewables: Costs and Markets to 2030. 2017, International Renewables Energy Agency: Abu Dhabi. 14. Belharouak, I., G.M. Koenig, and K. Amine, Electrochemistry and safety of Li4Ti5O12 and graphite anodes paired with LiMn2O4 for hybrid electric vehicle Li-ion battery applications. Journal of Power Sources, 2011. 196(23): p. 10344-10350. 15. Ashuri, M., Q. He, and L.L. Shaw, Silicon as a potential anode material for Li-ion batteries: where size, geometry and structure matter. Nanoscale, 2016. 8(1): p. 74-103. 16. Chan, C.K., et al., High-performance lithium battery anodes using silicon nanowires. Nature Nanotechnology, 2007. 3: p. 31. 17. Nitta, N., et al., Li-ion battery materials: present and future. Materials Today, 2015. 18(5): p. 252-264. 18. Obrovac, M.N. and L.J. Krause, Reversible Cycling of Crystalline Silicon Powder. Journal of The Electrochemical Society, 2007. 154(2): p. A103-A108. 19. Wu, H. and Y. Cui, Designing nanostructured Si anodes for high energy lithium ion batteries. Nano Today, 2012. 7(5): p. 414-429. 20. Choi, J.W. and D. Aurbach, Promise and reality of post-lithium-ion batteries with high energy densities. Nature Reviews Materials, 2016. 1: p. 16013. 21. Casimir, A., et al., Silicon-based anodes for lithium-ion batteries: Effectiveness of materials synthesis and electrode preparation. Nano Energy, 2016. 27: p. 359-376. 22. Wu, H., et al., Stable cycling of double-walled silicon nanotube battery anodes through solid electrolyte interphase control. Nature Nanotechnology, 2012. 7: p. 310. 23. Liu, N., et al., A pomegranate-inspired nanoscale design for large-volume-change lithium battery anodes. Nature Nanotechnology, 2014. 9: p. 187. 12

24. Chen, T., et al., Recent advancement of SiOx based anodes for lithium-ion batteries. Journal of Power Sources, 2017. 363: p. 126-144. 25. Yasuda, K., et al., Thermodynamic analysis and effect of crystallinity for silicon monoxide negative electrode for lithium ion batteries. Journal of Power Sources, 2016. 329: p. 462-472. 26. Hu, M., X. Pang, and Z. Zhou, Recent progress in high-voltage lithium ion batteries. Journal of Power Sources, 2013. 237: p. 229-242. 27. Li, W., B. Song, and A. Manthiram, High-voltage positive electrode materials for lithium-ion batteries. Chemical Society Reviews, 2017. 46(10): p. 3006-3059. 28. Liu, D., et al., Spinel materials for high-voltage cathodes in Li-ion batteries. RSC Advances, 2014. 4(1): p. 154-167. 29. Yang, S., W. Ren, and J. Chen, Li4SiO4-coated LiNi0.5Mn1.5O4 as the high performance cathode materials for lithium-ion batteries. Frontiers in Energy, 2017. 11(3): p. 374-382. 30. Zhou, L., D. Zhao, and X. Lou, LiNi0.5Mn1.5O4 Hollow Structures as High-Performance Cathodes for Lithium-Ion Batteries. Angewandte Chemie International Edition, 2012. 51(1): p. 239-241. 31. Schipper, F., et al., Review Recent Advances and Remaining Challenges for Lithium Ion Battery Cathodes: I. Nickel-Rich, LiNixCoyMnzO2. Journal of The Electrochemical Society, 2017. 164(1): p. A6220-A6228. 32. Liu, S., L. Xiong, and C. He, Long cycle life lithium ion battery with lithium nickel cobalt manganese oxide (NCM) cathode. Journal of Power Sources, 2014. 261: p. 285-291. 33. Myung, S.-T., et al., Nickel-Rich Layered Cathode Materials for Automotive Lithium-Ion Batteries: Achievements and Perspectives. ACS Energy Letters, 2017. 2(1): p. 196-223. 34. Tian, J., et al., High-Rate and Cycling-Stable Nickel-Rich Cathode Materials with Enhanced Li+ Diffusion Pathway. ACS Applied Materials & Interfaces, 2016. 8(1): p. 582-587. 35. Tang, Z., et al., Improvement of cycling and thermal stability of LiNi0.8Mn0.1Co0.1O2 cathode material by secondly treating process. Ionics, 2018. 24(1): p. 61-71. 36. Sun, H.-H. and A. Manthiram, Impact of Microcrack Generation and Surface Degradation on a Nickel-Rich Layered Li[Ni0.9Co0.05Mn0.05]O2 Cathode for Lithium-Ion Batteries. Chemistry of Materials, 2017. 29(19): p. 8486-8493. 37. Wu, F., et al., Effect of Ni2+ Content on Lithium/Nickel Disorder for Ni-Rich Cathode Materials. ACS Applied Materials & Interfaces, 2015. 7(14): p. 7702-7708. 38. Lee, K.-S., et al., Structural and Electrochemical Properties of Layered Li [ Ni1 2x Co x Mn x ] O2 ( x = 0.1 0.3 ) Positive Electrode Materials for Li-Ion Batteries. Journal of The Electrochemical Society, 2007. 154(10): p. A971-A977. 39. Kim, H., et al., A New Coating Method for Alleviating Surface Degradation of LiNi0.6Co0.2Mn0.2O2 Cathode Material: Nanoscale Surface Treatment of Primary Particles. Nano Letters, 2015. 15(3): p. 2111-2119. 40. Yang, J. and Y. Xia, Enhancement on the Cycling Stability of the Layered Ni-Rich Oxide Cathode by In-Situ Fabricating Nano-Thickness Cation-Mixing Layers. Journal of The Electrochemical Society, 2016. 163(13): p. A2665-A2672. 41. Lim, B.-B., et al., Comparative Study of Ni-Rich Layered Cathodes for Rechargeable Lithium Batteries: Li[Ni0.85Co0.11Al0.04]O2 and Li[Ni0.84Co0.06Mn0.09Al0.01]O2 with Two-Step Full Concentration Gradients. ACS Energy Letters, 2016. 1(1): p. 283-289. 42. Lim, B.-B., et al., Advanced Concentration Gradient Cathode Material with Two-Slope for High- Energy and Safe Lithium Batteries. Advanced Functional Materials, 2015. 25(29): p. 4673-4680. 43. Liu, B., et al., Improved electrochemical properties of Li1.2Ni0.18Mn0.59Co0.03O2 by surface modification with LiCoPO4. Electrochimica Acta, 2011. 56(19): p. 6748-6751. 44. Wang, Q.Y., et al., High capacity double-layer surface modified Li[Li0.2Mn0.54Ni0.13Co0.13]O2 cathode with improved rate capability. Journal of Materials Chemistry, 2009. 19(28): p. 4965-4972. 13

45. Qiao, Q.Q., et al., Surface modification of Li-rich layered Li(Li0.17Ni0.25Mn0.58)O2 oxide with Li-Mn-PO4 as the cathode for lithium-ion batteries. Journal of Materials Chemistry A, 2013. 1(17): p. 5262-5268. 46. Yang, X., et al., Spherical lithium-rich layered Li1.13[Mn0.534Ni0.233Co0.233]0.87O2 with concentration-gradient outer layer as high-performance cathodes for lithium ion batteries. Journal of Power Sources, 2013. 232: p. 338-347. 47. Wang, R., et al., Atomic Structure of Li2MnO3 after Partial Delithiation and Re-Lithiation. Advanced Energy Materials, 2013. 3(10): p. 1358-1367. 48. Rozier, P. and J.M. Tarascon, Review Li-Rich Layered Oxide Cathodes for Next-Generation Li- Ion Batteries: Chances and Challenges. Journal of The Electrochemical Society, 2015. 162(14): p. A2490-A2499. 49. Sathiya, M., et al., Origin of voltage decay in high-capacity layered oxide electrodes. Nature Materials, 2014. 14: p. 230. 50. Fang, R., et al., More Reliable Lithium Sulfur Batteries: Status, Solutions and Prospects. Advanced Materials, 2017. 29(48). 51. Matthey, J., Our Guide to Batteries - 3rd Edition. 2017. 52. Manthiram, A., S.-H. Chung, and C. Zu, Lithium Sulfur Batteries: Progress and Prospects. Advanced Materials, 2015. 27(12): p. 1980-2006. 53. Wang, D.-W., et al., Carbon-sulfur composites for Li-S batteries: status and prospects. Journal of Materials Chemistry A, 2013. 1(33): p. 9382-9394. 54. Ji, X., K.T. Lee, and L.F. Nazar, A highly ordered nanostructured carbon sulphur cathode for lithium sulphur batteries. Nature Materials, 2009. 8: p. 500. 55. Elazari, R., et al., Sulfur-Impregnated Activated Carbon Fiber Cloth as a Binder-Free Cathode for Rechargeable Li-S Batteries. Advanced Materials, 2011. 23(47): p. 5641-5644. 56. Borchardt, L., M. Oschatz, and S. Kaskel, Carbon Materials for Lithium Sulfur Batteries Ten Critical Questions. Chemistry A European Journal, 2016. 22(22): p. 7324-7351. 57. Cheng, X.-B., J.-Q. Huang, and Q. Zhang, Review Li Metal Anode in Working Lithium-Sulfur Batteries. Journal of The Electrochemical Society, 2018. 165(1): p. A6058-A6072. 58. Aurbach, D., et al., On the Surface Chemical Aspects of Very High Energy Density, Rechargeable Li Sulfur Batteries. Journal of The Electrochemical Society, 2009. 156(8): p. A694-A702. 59. Cheng, X.-B., et al., Toward Safe Lithium Metal Anode in Rechargeable Batteries: A Review. Chemical Reviews, 2017. 117(15): p. 10403-10473. 60. Li, W., et al., The synergetic effect of lithium polysulfide and lithium nitrate to prevent lithium dendrite growth. Nature Communications, 2015. 6: p. 7436. 61. Zhang, L., et al., The synergetic interaction between LiNO3 and lithium polysulfides for suppressing shuttle effect of lithium-sulfur batteries. Energy Storage Materials, 2018. 11: p. 24-29. 62. Zheng, G., et al., Interconnected hollow carbon nanospheres for stable lithium metal anodes. Nature Nanotechnology, 2014. 9: p. 618. 63. Balach, J., et al., Mesoporous Carbon Interlayers with Tailored Pore Volume as Polysulfide Reservoir for High-Energy Lithium Sulfur Batteries. The Journal of Physical Chemistry C, 2015. 119(9): p. 4580-4587. 64. Huang, J.-Q., Q. Zhang, and F. Wei, Multi-functional separator/interlayer system for highstable lithium-sulfur batteries: Progress and prospects. Energy Storage Materials, 2015. 1: p. 127-145. 65. Huang, C., et al., Manipulating surface reactions in lithium sulphur batteries using hybrid anode structures. Nature Communications, 2014. 5: p. 3015. 66. Ma, G., et al., A lithium anode protection guided highly-stable lithium-sulfur battery. Chemical Communications, 2014. 50(91): p. 14209-14212. 67. Cheng, X.-B., et al., The gap between long lifespan Li-S coin and pouch cells: The importance of lithium metal anode protection. Energy Storage Materials, 2017. 6: p. 18-25. 14

68. Rahman, M.A., et al., High Energy Density Metal-Air Batteries: A Review. Journal of the Electrochemical Society, 2013. 160: p. A1759-A1771. 69. Ma, Z., et al., A review of cathode materials and structures for rechargeable lithium-air batteries, Energy & Environmental Science, 2015. 8: p. 2144-2198. 70. Manthiram, A., X. Yu, and S. Wang, Lithium battery chemistries enabled by solid-state electrolytes. Nature Reviews Materials, 2017. 2: p. 16103. 71. Bachman, J.C., et al., Inorganic Solid-State Electrolytes for Lithium Batteries: Mechanisms and Properties Governing Ion Conduction. Chemical Reviews, 2016. 116(1): p. 140-162. 72. Yue, L., et al., All solid-state polymer electrolytes for high-performance lithium ion batteries. Energy Storage Materials, 2016. 5: p. 139-164. 73. Li, J., et al., Solid Electrolyte: the Key for High-Voltage Lithium Batteries. Advanced Energy Materials, 2015. 5(4): p. 1401408-n/a. 74. Slater, M.D., et al., Sodium-Ion Batteries, Advanced Funcional Materials, 2013. 23: p. 947-958. 75. Yabuuchi, N., et al., P2-type Na x [Fe 1/2 Mn 1/2] O 2 made from earth-abundant elements for rechargeable Na batteries, Nature materials, 2012, 11: p. 512-5127. 76. Xia, X., et al., Study of the Reactivity of the Na/Hard Carbon with Different Solvents and Electrolytes, Journal of the Electrochemical Society, 2012, 159: p. A515-A519. 77. Deng, J., et al., Sodium-Ion Batteries: From Academic Research to Practical Commercialization, Advanced Energy Materials, 2017, p. 1701428(1)-1701428(17). 78. Pan, H., et al., Room-temperature stationary sodium-ion batteries for large-scale electric energy storage, Energy & Environmental Science, 2013, 6: 2338-2360. 79. Yabuuchi, K., et al., Research Devlopment on Sodium-Ion Batteries, Chemical Reviews, 2014, 114: p. 11636-11682. 80. Komaba, S., et al., Study on the Reversible Electrode Reaction of Na 1-xNi 0.5Mn 0.5O 2for a Rechargeable Sodium-Ion Battery, 2012, 51: p. 6211-6220. 81. Choi, J.W., et al., Promise and reality of post-lithium-ion batteries with high energy densities, 2016, 1: p. 1-16. 82. Langridge, M. and Edwards, L., Future batteries, coming soon: Charge in seconds, last months and power over the air, https://www.pocket-lint.com/gadgets/news/130380-futurebatteries-coming-soon-charge-in-seconds-last-months-and-power-over-the-air, 2017, (accessed: January 2018). 83. Massé, R.C., et al., Beyond Li-ion: electrode materials for sodium- and magensium-ion batteries, Science China Materials, 2015, 58: p. 715-766. 84. Lin, M.-C., et al., An ultrafast rechargeable aluminium-ion battery, Nature, 2015, 520: p. 325-328. 85. Faradion Limited, Sodium-Ion Technology, http://www.faradion.co.uk/, (accessed: January 2018). 86. OXIS Energy, Next Genergation Battery Technology, https://oxisenergy.com/, (accessed: Janaury 2018). 15