Rechargeable Li-ion batteries

The introduction of non-aqueous rechargeable Li-ion batteries in the 1970s and the commercialization in the 1990s to power portable electronic devices, such as cellular phones and laptop computers, sparked a revolution in battery technology. This marked a massive swing away from the relatively low-voltage, water-based systems such as Ni-Cd and nickel metal hydride (NiMH) batteries as well as high-temperature systems, such as conventional Na-S batteries. Nowadays, rechargeable Li-ion batteries are being pursued intensively for a myriad of devices, such as uninterrupted power supply units, rechargeable power sources for consumer electronics, and electrical vehicles.

The first commercial Li-ion battery, introduced by Sony Corporation in 1991, was based on a LiCoO₂ cathode and a carbon anode, as schematically shown in Figure 1. When an electrical current is applied to charge the cell, lithium ions move out of the cathode (Li_1–xCoO₂) and become trapped inside the anode storage medium, which is usually graphitized carbon (Li_xC₆). Upon battery discharge, the lithium ions travel back to the cathode and produce an external electrical current. During cell operation at 3.0–4.2 V, however, the surface reactivity and instability of the delithiated Li_1–xCoO₂ structure limit the practical capacity of the LiCoO₂ electrodes to approximately 140 mAh/g, which corresponds to x ≈ 0.5 (i.e., ∼50% of its theoretical value [273 mAh/g]).^{Reference Lee, Carlton, Risch, Surendranath, Chen, Furutsuki, Yamada, Nocera and Shao-Horn10} These limitations, together with the high possibility of thermal runaway caused by cell overcharge and short circuit in inadequately controlled batteries and the relatively high cost of cobalt, have led to enormous efforts since 1991 to find alternative cathode materials to LiCoO₂ that provide Li-ion cells with superior energy density, rate capability, safety, and cycle life.

Figure 1. Scheme of a common lithium-ion battery and its electrochemical reaction. Typically, a rechargeable Li-ion battery consists of two Li-ion intercalation electrodes, for instance, a graphite anode and a layered LiCoO₂ cathode, with a non-aqueous electrolyte in between for ionic conduction. The electric and chemical energies in a Li-ion cell are interconverted through reversible discharge/charge processes between the cathode and anode along with electrons traveling through an external circuit simultaneously. The overall electrochemical reaction for the C/LiCoO₂ cell is given on the right side, where φ⁰ represents the standard redox potential of the electrodes, and E ⁰ represents the cell voltage, respectively.

Insertion electrodes for Li-ion electrochemical cells need to have stable structures over a wide compositional range such that as much lithium as possible can be extracted and reinserted during repeated charging and discharging to maximize cell energy density and cycle life.^{Reference Whittingham1} Furthermore, the host structures must have interstitial spaces that provide energetically favorable pathways for fast Li-ion transport—that is, high power capability. Since carbon in the form of graphite is the material of choice for the anode for the Li-ion battery industry, these requirements must be achieved by varying the cathode materials. Incidentally, the nearly universal use of graphite anodes also explains why battery cell characteristics can be discussed in terms of the cathode’s properties alone.

Several alternative cathode materials to LiCoO₂ have been exploited by the Li-ion battery industry over the past decade. They include compositional variations of the layered LiCoO₂ structure, such as LiNi_0.8Co_0.15Al_0.05O₂ (NCA);^{Reference Bang, Joachin, Yang, Amine and Prakash11} spinel electrodes derived from LiMn₂O₄, such as lithium-rich compounds in the Li_1+xMn_2–xO₄ system;^{Reference Gu, Belharouak, Zheng, Wu, Xiao, Genc, Amine, Thevuthasan, Baer, Zhang, Browning, Liu and Wang12} and LiFePO₄ that has an olivine-type structure.^{Reference Padhi, Nanjundaswamy, Masquelier, Okada and Goodenough13} Although NCA provides a slightly higher practical capacity (160–180 mAh/g) than LiCoO₂, its thermal instability on delithiation due to the presence of the high valence Ni compromises the safety of Li-ion cells. On the other hand, spinel LiMn₂O₄ and olivine LiFePO₄ electrodes are significantly more stable to lithium extraction than the layered Co- and Ni-based electrodes (both structurally and thermally), but they deliver relatively low practical capacities in a lithium cell above 3 V, typically 100–150 mAh/g at moderate current rates.

It became clear by the end of the 1990s that new strategies would have to be developed to design alternative high potential cathode materials (>3 V) with capacities superior to those achievable with standard LiCoO₂-, LiMn₂O₄-, and LiFePO₄-type electrodes without compromising the structural stability or rate capability of the electrodes or the cycle life of the cells. Recently, researches at Argonne National Laboratory have developed a family of high-energy manganese-based cathodes by structurally integrating a Li₂MnO₃ stabilizing component into an electrochemically active LiMO₂ (M = Mn, Ni, Co) electrode.^{Reference Johnson, Li, Lefief, Vaughey and Thackeray14–Reference Sun, Myung, Park, Prakash, Belharouak and Amine16} The relatively high Mn content in these high-energy cathode materials lowers material costs, while the excess lithium boosts specific capacity up to 250 mAh/g between 4.6 and 2.5 V, and therefore, significantly improves the energy density of battery cells to 900 Wh/kg. However, in practical cells, when these high-energy Ni-Mn-Co oxides (NMC) are cycled against graphite, deliverable capacity decreases dramatically with cycle number along with a significant decay of cell discharge voltage.^{Reference Li, Bettge, Polzin, Zhu, Balasubramanian and Abraham17} The consequences of the capacity and voltage fading of these materials is the severe loss of the energy density of the cell during long cycling times, which hinders its practical application in electrical vehicles. The underlying mechanisms for the observed energy fading need to be addressed in order to unlock the potential of these compounds as high-energy cathode materials for Li-ion batteries.

Furthermore, since these cathodes operate at a high voltage, there is a need to develop high-voltage electrolytes to enable these new cathodes.^{Reference Abouimrane, Odom, Tavassol, Schulmerich, Wu, Bhargava, Gewirth, Moore and Amine18,Reference Abouimrane, Belharouak and Amine19} Several novel organic solvents with greater oxidative stability, such as sulfones,^{Reference Abouimrane, Belharouak and Amine19} nitriles,^{Reference Nagahama, Hasegawa and Okada20} and fluorinated solvents,^{Reference Hu, Zhang and Amine21} have recently been explored as electrolytes. Unfortunately, these electrolytes may also compromise the anode solid electrolyte interphase-forming reactions required in Li-ion batteries. The development of a novel electrolyte additive that helps form an interfacial film on the cathode surface is thus important and will likely lead to the development of workable electrolyte systems for high-voltage cathodes.^{Reference Abouimrane, Odom, Tavassol, Schulmerich, Wu, Bhargava, Gewirth, Moore and Amine18}

Since lithium resources are not considered to be abundant, there is the potential for significant cost increase if vehicle electrification expands in the future. As a result, there is growing interest in substituting lithium ions with sodium ions, because sodium is one of the most abundant elements. The article by Kubota et al. in this issue addresses progress in and challenges of sodium ion batteries.

Anode materials for rechargeable Li-ion batteries

Li batteries with metallic Li anodes offer one of the highest theoretical capacities among conventional battery types, and, in principle, should provide the highest energy density of all Li batteries, primary or secondary, since lithium metal has an extremely high specific capacity (3860 mAh/g) and lower negative redox potential (–3.04 V versus standard hydrogen electrode [SHE]).^{Reference Aurbach and Cohen22} However, two major technical bottlenecks prevent the realization of a successful rechargeable Li metal battery.^{Reference Xu, Wang, Ding, Chen, Nasybulin, Zhang and Zhang23} One is the growth of lithium dendrites during repeated charge/discharge cycles, which severely compromises the rechargeability of each lithium cell. The rechargeability is affected by the reactions that can take place between the nonaqueous, flammable electrolyte and the cycled lithium anode, leading to the formation of high surface area dendrites. The formation of the lithium dendrites could also lead to serious safety hazards because of the potential for internal short circuits if these dendrites penetrate through the separators and contact the cathode directly. The other bottleneck is low Coulombic efficiency during repeated cycles, although this can be partially compensated for by an excess amount of lithium. For example, in the early development of Li metal batteries, an excess amount of 300% of lithium was typically applied. Overcoming these hurdles presents an enormous challenge to the lithium battery industry.

Recently, researchers demonstrated that the growth of lithium dendrites can be partially prevented through either a physical blocking mechanism (using polyethylene oxide-based block copolymer electrolytes)^{Reference Balsara, Singh, Eitouni and Gomez24} or a self-healing mechanism (using electrolyte additives).^{Reference Ding, Xu, Graff, Zhang, Sushko, Chen, Shao, Engelhard, Nie, Xiao, Liu, Sushko, Liu and Zhang25} However, these mechanisms are only effective under very limited conditions (i.e., at high temperatures or under low current densities). Therefore, more work is needed to explore a more reliable solution to prevent dendrite growth in order to push the use of lithium anodes for broader applications. Despite these obstacles, significant efforts are under way to capitalize on and exploit the advantages of metallic lithium systems, such as Li-S and Li-air batteries (see the Nazar et al. and Kwabi et al. articles, respectively, in this issue), with a big assumption that these obstacles can be overcome eventually.

The technical hurdles of lithium metal as the anode material have led to the use of carbon-based materials as the most widely used negative electrodes in current rechargeable Li-ion batteries, typically carbon-based materials with the limited specific capacity of graphite (372 mAh/g). In order to overcome the capacity limits of current technology, materials such as Sn, Sb, Si, and Ge,^{Reference Terranova, Orlanducci, Tamburri, Guglielmotti and Rossi8,Reference Wang, Li, Qiu, Luo, Ning, Li, Zhang, Liang and Zhi26–Reference Cho and Picraux28} which form alloys with lithium, have been explored as potentially more attractive anode candidates since they can incorporate larger amounts of lithium (Figure 2, Sn and Si are shown because only they have been intensively investigated thus far). Among these, silicon-based anodes are particularly attractive because of their higher theoretical specific capacity of approximately 4200 mAh/g (ca. Li_4.4Si), which is far larger than that of graphite and oxide materials.^{Reference Ge, Fang, Rong and Zhou27} However, the application of bulk silicon anodes faces one major problem: during the reaction for formation of the silicon-lithium alloy (corresponding to the insertion of lithium in the negative electrode during the charging process), volume expansion from the delithiated phase to the lithiated phase may reach 380%. This high expansion, followed by contraction of the same amplitude (corresponding to the extraction of lithium from the negative electrode during the discharging process), rapidly leads to irreversible mechanical damage to the electrode and eventually to a loss of contact between the negative electrode and the underlying current collector, which causes rapid capacity fade during cycling. Furthermore, silicon usually possesses low electrical conductivity, which has the effect of kinetically limiting the use of the battery. A significant effort is currently under way to enable this system by designing conductive binders that can minimize particle isolation or by incorporating Si in graphene sheets to maintain good conductivity at the electrode level during cycling.^{Reference Wang, Li, Qiu, Luo, Ning, Li, Zhang, Liang and Zhi26,Reference Ye, Xie, Rick, Chang and Hwang29}

Figure 2. Specific capacities of different anodes showing that silicon-based anodes are particularly attractive because of their higher theoretical specific capacity of approximately 4200 mAh/g (ca. Li_4.4Si), which is far larger than those of graphite and other alloy materials.

Beyond Li-ion systems

The inherent energy densities of current Li-ion technology are not sufficient for the long-term needs of future applications such as extended-range electrical vehicles. Going beyond Li-ion requires the exploration of new electrochemistries and materials, offering a great opportunity to reach the ultimate goal, although this represents a formidable challenge. In this section, we provide a brief overview of three such systems, rechargeable Li-S, Li-air, and Mg batteries, and we address some of the key challenges for each of these individual systems.

Li-S batteries

The rechargeable Li-S cell operates by reduction of S at the cathode upon discharge to form a series of soluble polysulfide species (Li₂S₈, Li₂S₆, Li₂S₄) that combine with Li to ultimately produce solid Li₂S₂ and Li₂S at the end of the discharge, as illustrated in Figure 3. On charging, Li₂S₂/Li₂S is converted back to S via similar soluble polysulfide intermediates presented in the discharge process and lithium plates to the nominal anode, making the cell reversible. This contrasts with conventional Li-ion cells, where the lithium ions are intercalated in the anode and cathode, and consequently the Li-S system, which allows for a much higher lithium storage density.^{Reference Yang, Zheng and Cui30,Reference Barghamadi, Kapoor and Wen31}

Figure 3. Scheme of a Li-S cell and its electrochemical reactions. The rechargeable Li-S cell operates by reduction of S at the cathode on discharge to form a series of soluble polysulfide species (Li₂S₈, Li₂S₆, Li₂S₄) that combine with Li to ultimately produce solid Li₂S₂ and Li₂S at the end of the discharge, with the process being reversed on charge, as shown on the left side of the figure. The overall reaction and a typical discharge profile of a Li-S cell are provided on the right side of the figure.

The Li-S batteries, when based on the overall reaction S₈ + 16 Li = 8 Li₂S, operate at an average voltage of 2.15 V with a theoretical specific capacity of 1675 mAh/g-S. This leads to an energy density of 2600 Wh/kg (2800 Wh/L), which is five times higher than that of the conventional Li-ion battery based on intercalation compounds. Sulfur is an abundant material available on a large scale and at low cost as a side product of petroleum and mineral refining, which makes it attractive for low-cost and high-energy rechargeable lithium batteries. Furthermore, the unique features of Li-S chemistry provide inherent chemical overcharge protection, which enhances safety, particularly for high-capacity multi-cell battery packs.^{Reference Yang, Zheng and Cui30}

Although sulfur-based electrochemical cells had been reported in 1962, initial drawbacks in terms of the electronically insulating nature of sulfur, the solubility of intermediately formed polysulfides in common liquid organic electrolytes, as well as widely known dendrite formation issues accompanying the use of metallic lithium as a negative electrode could not be overcome for several decades and are, in fact, still not solved satisfactorily. In addition, the formed polysulfides in the electrolyte migrate to the lithium metal anode and are electrochemically reduced, well known as a “shuttle reaction,”^{Reference Yan, Kai, Shizhao and Xiaobin32} which results in low Coulombic efficiency and rapid capacity fade in Li-S batteries.

Recently, interest in Li-S based rechargeable batteries has been steadily increasing thanks to the opportunities to design new nanostructured material architectures,^{Reference Yang, Zheng and Cui33–Reference Kim, Lim, Kim and Kang35} which could overcome issues related to the bulk material’s conductivity. Moreover, the development of new electrolytes, binder materials, and cell design concepts in general has led to significant advances in the field of Li-S based secondary batteries within the last few years.^{Reference Barghamadi, Kapoor and Wen31} There is no doubt that Li-S batteries will remain attractive over the longer term because of their inherently high-energy content, high power capability, and potential for low cost, although they are still in the development stage (see the Nazar et al. article in this issue).

Li-air batteries

Li-air batteries offer superior theoretical energy density and are considered to be the “holy grail” of lithium batteries (Table I). The energy density of Li-air batteries is over an order of magnitude higher than Li-ion batteries. Whereas state-of-the-art Li-ion batteries have achieved 150–200 Wh/kg (of the 900 Wh/kg theoretically possible value) at the cell level, Li-air batteries have the potential to achieve 3620 Wh/kg (when discharged to Li₂O₂ at 3.1 V) or 5200 Wh/kg (when discharged to Li₂O at 3.1 V). When the “free” oxygen supplied during discharge and released during charge is not included in the calculation, Li-air cells offer ∼11,000 Wh/kg. This is basically identical to the value for gasoline (octane) at ∼13,000 Wh/kg when the oxygen that is supplied externally, combusted within, and exhausted from the engine is neglected. Unlike other battery technologies, Li-air is thus competitive with liquid fuels.

Table I. Relative specific energies of Li/O₂, Li/S, Li-ion, and gasoline systems.

Note: OCV, open-circuit voltage

During discharge of the Li-air cell, Li is oxidized to Li⁺ at a metallic Li anode, which conducts through an electrolyte composed of a non-aqueous solvent and a Li salt, and reacts with O₂ from air on a cathode composed of carbon, a catalyst, and a binder deposited on a carbon paper substrate, as shown in Figure 4. The Li-air technology has the potential to significantly reduce the cost well below that of the Li-ion battery due to the higher specific energy densities and the lower cost of the proposed cell components, in particular of the carbon-based cathode materials versus the nickel, manganese, cobalt oxides used in Li-ion battery cathodes.^{Reference Lu, Gallant, Kwabi, Harding, Mitchell, Whittingham and Shao-Horn36–Reference Lu and Amine38} A non-aqueous electrolyte is preferred, as it has been shown to have higher theoretical energy densities than aqueous electrolyte designs.^{Reference Zheng, Liang, Hendrickson and Plichta39}

Figure 4. Diagram of a non-aqueous Li-air battery. A typical non-aqueous Li-air cell is composed of a lithium electrode, an electrolyte consisting of dissolved lithium salt in an organic solvent, and a porous O₂-breathing electrode that contains carbon particles and, in some cases, an added electrocatalyst.

Current Li-air batteries are still in the experimental stages, and the realization of the high theoretical energy densities and practical application of this technology have been limited by the low power output (i.e., low current density), poor cycleability, and low energy efficiency of the cell. These limitations are caused by the materials and system design:

1. (1) Unstable electrolytes.^{Reference McCloskey, Scheffler, Speidel, Bethune, Shelby and Luntz40,Reference Freunberger, Chen, Peng, Griffin, Hardwick, Barde, Novak and Bruce41} The current non-aqueous, carbonate electrolytes are volatile, unstable at high potentials, easily oxidized, and reduced at the lithium anode in the presence of crossover oxygen. This seriously limits cycle life.

2. (2) Lithium electrode poisoning due to oxygen crossover and reaction with the electrolyte destroys the integrity and functioning of the cell.^{Reference Assary, Lu, Du, Luo , Zhang, Ren, Curtiss and Amine42} This also lowers cycle life.

3. (3) Li₂O₂ and/or Li₂O deposition on the carbon cathode surface or within the pores creates clogging and restricts the oxygen flow.^{Reference Lu and Shao-Horn43,Reference Lu, Kwabi, Yao, Harding, Zhou, Zuin and Shao-Horn44} This lowers capacity.

4. (4) Inefficient cathode structure and catalysis.^{Reference Shao, Ding, Xiao, Zhang, Xu, Park, Zhang, Wang and Liu45–Reference Shao, Park, Xiao, Zhang, Wang and Liu47} Commonly used carbons and cathode catalysts do not access the full capacity of the oxygen electrode and cause significant charge overpotentials. This lowers the power capability.

It has recently become apparent that the electrolyte plays a key role in the Li-air cell performance.^{Reference Black, Adams and Nazar48–Reference Jung, Hassoun, Park, Sun and Scrosati50} The oxygen anion radical O₂^– intermediate or other reduction species, which may form during the discharge process, can be highly reactive and may cause the electrochemical response to be dominated by electrolyte decomposition rather than the expected lithium peroxide formation. Developing a stable electrolyte^{Reference Zhang, Lu, Assary, Du, Wang, Sun, Qin, Lau, Greeley, Redfern, Iddir, Curtiss and Amine51} as well as the materials and their microstructures in the O₂-breathing cathode^{Reference Trahey, Karan, Chan, Lu, Ren, Greeley, Balasubramanian, Burrell, Curtiss and Thackeray52–Reference Lei, Lu, Luo, Wu, Du, Zhang, Ren, Miller, Sun and Elam58} will certainly advance Li-air technology close to application. For further insights into Li-air batteries, see the article by Kwabi et al. in this issue.

In this issue

Croy et al. present a detailed description of next-generation Li-ion batteries by including near-term advancements in high-energy lithium-metal-oxide cathode materials, high-energy alloy and oxide anode materials, and high-voltage fluorine-based electrolytes. These high energy and high voltage chemistries are expected to be used in future all-electric vehicles. Furthermore, there is increasing interest worldwide in developing low cost and sustainable systems for grid energy storage that does not use lithium. Rechargeable Na-ion batteries, due to the almost infinite supply of Na, are the most appealing as an immediate alternative to lithium batteries. In this issue, Kubota et al. and Yamada describe the progress, challenges, and future directions for rechargeable Na-ion batteries, with particular focus on the layered oxide (Kubota et al.) and iron-based (Yamada) cathode materials.

Lithium metal could be an ideal anode for next-generation high-energy rechargeable batteries, including Li-sulfur and Li-air batteries. However, two major technical bottlenecks prevent the realization of a successful rechargeable Li metal battery (i.e., the growth of dendrites and low Columbic efficiency). Vaughey et al. describe several approaches to stabilize the surface of lithium metal and minimize the dendritic growth. This article will also offer a detailed description of the technologies beyond Li-ion, such as Li-sulfur and Li-air batteries, as reviewed by Nazar et al. and Kwabi et al., respectively. Enabling technologies beyond lithium ion will lead to significant cost reduction and an increase in the electrical driving range, leading to an expansion in the electrification of vehicles. The final article of this issue by Shterenberg et al. details the challenge of developing rechargeable magnesium batteries.

Concluding remarks

This overview article presents a brief glimpse into rechargeable Li/Mg batteries as energy storage and conversion devices for electrical vehicle and grid applications. These battery systems include rechargeable Li-ion batteries with specific emphasis on high-energy cathode based Li-ion cells, rechargeable Li-S, Li-air, and Mg batteries, which possess the electrochemistries that go beyond the conventional intercalated Li-ion systems. Future requirements for batteries demand innovative concepts for charge storage at the interface of the electrode and electrolyte. These concepts will be realized only by gaining a fundamental understanding of the chemical and physical processes that occur at this complex interface.

Our hope is that the concepts and results presented in this issue of MRS Bulletin will prompt new researchers to join this field and help broaden the scope and impact of rechargeable Li/Mg batteries. While valuable progress has been achieved over the past decades, we believe that the most significant advancements and impacts still await discovery and understanding, which could help realize a true transition to an electrified transportation system in the near future.