Lithium Batteries

Abstract

First attempts to create batteries using an ion other than the proton were done in the 1970s with the fabrication of lithium primary cells. It was the fast development of the electronic devices that pouch electrochemists in the new world of lithium. After primary cells came secondary (rechargeable) lithium batteries in the 1980s. Innovations and advances in insertion electrode materials have improved the stored energy compared with other systems. For half-a-century, lithium batteries are increasingly used in a huge number of applications from watches, portable electronics to electric transportation and stationary grid storage. While older technologies such as Zn-MnO₂, lead-acid, and Ni-Cd are still used, the increasing battery market is now dominated by Li-ion batteries. The purpose of this chapter is to introduce the technologies of primary and secondary lithium electrochemical cells with a special focus on lithium-ion batteries and lithium-metal polymer batteries.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

2.1 Introduction

By placing a very pure lithium-metal foil as anode element and a lithium salt in a nonaqueous solution as electrolyte, a new generation of electrochemical generators was born in the mid-1960s. Basically, the charge transport is identical to nickel-metal hydride (Ni-MH) or nickel-cadmium (Ni-Cd) batteries, except that Li⁺ ions are created by the simple reaction

$$ \mathrm{L}\mathrm{i}\to {\mathrm{Li}}^{+}+{e}^{-}, $$

(2.1)

liberating one electron through the external circuit and one ion introduced into the porous structure of the cathode. As discussed in the first chapter, the main parameters of energy storage systems are energy density (gravimetric and volumetric), power density, energy efficiency, and energy quality [1]. The great attractiveness of the lithium technology comes from the fact that Li is a lighter metal (molar weight M _w = 6.941 g mol⁻¹ and density 0.51 g cm⁻³), with the electronic configuration (He)2s¹. The specific capacity of Li metal is 3860 mAh g⁻¹ and the couple Li⁰ has the highest electroactivity with standard redox potential −3.04 V against H₂/H⁺. Consequently, the voltage of lithium batteries is also significantly higher than that of the Pb-acid and Ni-metal hydride, because lithium is the most electropositive element found in nature.

Primary and secondary lithium batteries using a nonaqueous electrolyte , exhibit higher energy density than aqueous electrolyte-based batteries due to the cell potential higher than 1.23 V, the thermodynamic limitation of water at 25 °C. The excellent performances of nonaqueous lithium batteries may meet the need for high power batteries in micro-devices, portable equipment, and even electrical vehicles.

Lithium does not exist in the form of pure metal in nature due to a very high reactivity with air, nitrogen and water. It is extracted from ore or brine salt marsh (lithium chloride LiCl, lithium hydroxide LiOH, lithium carbonate Li₂CO₃). Since we only have 112.7 g of lithium in one kilo of Li₂CO₃, the extraction of one kilogram of lithium requires 5.3 kg of lithium carbonate. According to the US Geological survey (January 2010), Bolivia would house 32 % of world reserves of lithium carbonate (Li₂CO₃), and Chile nearly 27 %. However, Chile, China, and Argentina are the largest producers [2]. Note that 0.8 kg of lithium metal is produced per second in the world that is 25000 tons a year, mainly to produce lithium-ion batteries for electric cars or cell phones. In the US Geologican survey, Goonan said that it would take 1.4–3.0 kg of lithium equivalent (7.5–16.0 kg of lithium carbonate) to support a 40-mile trip in an electric vehicle before requiring recharge [2].

In this chapter, attention is focused on the technology of the lithium batteries. After a brief historical overview, we draw most of the primary and secondary lithium batteries. To get further information on the interest of these power sources, we invite the readers to consult works in references as there have been published numerous excellent books on LiBs based on various different viewpoints [3–16]. However, there are few books available on the state-of-the-art and future of next generation LiBs, particularly eventually for EVs and HEVs.

2.2 Historical Overview

Research in lithium batteries began in 1912 under G.N. Lewis, but the breakthrough came in 1958 when Harris noticed the stability of Li-metal in a number of nonaqueous (aprotic) electrolytes such as fused salts, liquid SO₂, or lithium salt into an organic solvent such as LiClO₄ in propylene carbonate (C₄H₆O₃). The formation of a passivation layer that prevents the direct chemical reaction between lithium metal and the electrolyte but still allows for ionic transport is at the origin of the stability of lithium batteries [17].

These researches opened the door to the fabrication and commercialization of varieties of primary lithium batteries; since the late 1960s nonaqueous lithium cells, especially the 3-V primary systems, have been developed. These systems include lithium-sulfur dioxide (Li//SO₂) cells, lithium-polycarbon monofluoride (Li//(CF _x ) _n ) cells introduced by Matsuschita in 1973, lithium-manganese oxide (Li//MnO₂) cells commercialized by Sanyo in 1975, lithium-copper oxide (Li//CuO) cells, lithium-iodine (Li//(P2VP)I _n ) cells. During the same period, molten salt systems (LiCl-KCl eutecticum) using a Li-Al alloy anode and a FeS cathode were introduced [1]. The lithium-iodine battery has been used to power more than four million cardiac pacemakers since its introduction in 1972. During this time the lithium-iodine system has established a record of reliability and performance unsurpassed by any other electrochemical power source [18].

In the early 1970s was discovered the reversible insertion of guest species (ions, organic molecules, organometallics) into an host lattice that maintains its structural features but exhibits new physical properties. The first work was initiated by Armand at Stanford [19] on the properties of Prussian blue (iron cyanide bronzes M_0.5Fe(CN)₃). Among a great variety of inorganic materials, transition-metal oxide (TMOs) as CrO₃ was investigated by Armand at Grenoble [20] and transition-metal dichalcogenides (TMDs) were studied by DiSalvo at Bell Laboratories [21], and later by researchers at EXXON [22–25]. Rechargeable lithium cells using Li-insertion compounds as positive electrodes were developed in the mid 1970s, when Winn and Steele [26, 27] at Chloride Technical Ltd reported solid-solution electrode of TiS₂ with lithium and sodium as well, followed by the announcement of Exxon (USA) for its intention to commercialize the Li//TiS₂ system [28]; Bell Laboratories (USA) reported the development of lithium cells constructed with either NbSe₃ or TiS₃ [29], while transition-metal oxides such as V₂O₅, V₆O₁₃ were investigated by Dickens in the UK [30]. Table 2.1 documents the numerous efforts toward the development of rechargeable lithium batteries using TMOs as positive electrode and Li metal as anode.

Table 2.1 Rechargeable lithium metal batteries using TM oxides as positive electrodes

Prototype cells using the principle of lithium intercalation showed promising characteristics of very high energy density (for instance, 3 times that of Ni-MH cells), moderate cycle life and higher voltage (~3 V) than aqueous technologies. Beside the development of positive electrodes, Li alloys (for instance, Li-Al) were envisaged as negative electrodes. Among them, the first commercially available rechargeable lithium power source system, the Li//MoS₂ type battery (MOLICEL™) was manufactured by Moli Energy Ltd. in Canada. This cell delivers performance with sustained drain rates of several amperes at a cell voltage between 2.3 and 1.3 V. The energy density was in the 60–65 Wh kg⁻¹ range at a discharge rate of C/3 (approximately 800 mA). The total capacity of a C-size cell was 3.7 Ah and the number of realizable cycles is dependent upon the charge and discharge conditions [10]. Especially manufactured as the power source for pocket telephones in Japan, the first attempt to fabricate rechargeable batteries has proven very difficult ensuing safety concerns including fires due to the formation of dendrites. All the products were recalled [31].

At Eveready Battery Co. (USA) a Li/chalcogenide glass/TiS₂ was designed for the CMOS memory backup market. This cell is based on phosphorous chalcogenide glasses Li₈P₄O_0.25S_13.75 mixed with LiI and on solid-solution composite electrode TiS₂-solid electrolyte-black carbon with the percentage by weight of 51:42:7. The cathode capacity ranges from 1.0 to 9.5 mAh. The cell packaging is a standard sized XR2016 coin cell. The impedance of the cell at 21 °C is between 25 and 100 Ω depending upon where the cell is in the charge–discharge cycle. More than 200 cycles have been obtained in experimental cells [32]. A spirally wound AA-size Li//TiS₂ cell has been constructed by Grace Co. (USA). At 200 mA discharge rate, 1 Ah is delivered to 1.7 V [33]. In C-size, a Li//TiS₂ cell built by EIC Laboratories Inc. (USA), a capacity of 1.6 Ah was obtained. This cell operates in the temperature range from −20 to +20 °C [34].

The concept of lithium rocking-chair cells or lithium-ion batteries (LiBs) is not new. It has been proposed in the late 1970s by Armand who suggested to use two different intercalation compounds as a positive and negative electrode, in the so-called rocking-chair battery, the lithium ions being transferred from one side to the other [35]. Practically they were demonstrated in the early 1980s [36].

However, this concept has gained renewed attention following the success of Japanese industries. Sony and Sanyo in 1985 and 1988, respectively [37, 38], started to apply new electrodes materials following the fundamental researches of Goodenough in Oxford [39, 40] and of Armand and Touzain in Grenoble [41], who evidenced the fast motion of lithium ions in layered host structures (named as lithium intercalation compounds ). A LiB contains a high-voltage (cathode) and a low voltage (anode) host versus Li⁰/Li⁺ redox couple. Exploiting the idea for large-scale production, Sony Energytec Inc. has commercialized in June 1991 a Li-ion battery including a lithium cobaltate (LiCoO₂) and a non-graphitic carbon (lithiated coke LiC₆), as cathode and anode, respectively; target was powering mobile phones. The Li-ion battery by Sony was an 18650-type (253 Wh L⁻¹ energy density) using LiPF₆ in propylene carbonate/diethyl carbonate electrolyte solution [42]. Table 2.2 lists the initial applied patents for the invention of LiBs. The first patent related to the construction of a LiB, issued on 5 October 1985, is the property of the Japanese company Asahi Chemical Ind.

Table 2.2 Prior patents related to Li-ion batteries

During the following decades, active R&D was directed towards alternative electrode materials resulting in small, light and high energy density batteries that have subsequently allowed the rapid development of small portable electronics. Li-ion cells have been marketed later (Table 2.3) by several battery companies in the world (Sanyo, Matsushita, Hitachi, Yuasa, Moli, A&T, SAFT, etc.). Lithium-ion chemistries represent a significant step forward in battery technology. As an example, designed for sophisticated portable applications, the VL18650 from SAFT delivers a rated capacity of 1.2 Ah within stable operating conditions of discharge temperature in the range between −20 and +60 °C. Now, scientists and engineers have to throw down the challenge of developing high power LiBs for electric vehicles. Different varieties of Li-ion batteries were launched to power EV cars: in 1995, the 100-Ah LiCoO₂//graphite battery from Sony Corp. supplied energy for traveling 200 km at maximum speed up to 120 km h⁻¹; in 1996, the car launched by Mitsubishi Motors used LiMn₂O₄ spinel as cathode powering a trip of 250 km.

Table 2.3 Energy densities at a C/5 discharge rate of some commercialized Li-ion batteries

2.3 Primary Lithium Batteries

Developed in the early 1970s, primary lithium batteries are the most energy-dense electrochemical cells made for watches, film cameras, medical devices, and military purposes. Lithium primary cells have a typical gravimetric density 250 Wh kg⁻¹, against only 150 Wh kg⁻¹ for Li-ion batteries. Various technologies that differ in chemistry and construction have been used to develop primary lithium batteries. In Table 2.4, they are classified into three groups, according to the form and the type of cathode and electrolyte used. Frost and Sullivan said that in 2009, primary batteries including alkaline that were leading the market, carbon-zinc and lithium cells made 23.6 % of the global market with 3 % for primary lithium cells; it has been predicted a 7.4 % decline in 2015 due to the development of rechargeable batteries [43]. According the Freedonia’s studies, the production of primary batteries doubled between 2002 and 2012. US demand for primary and secondary batteries are expected to grow 4.2 % per year to $17.1 billion in 2017. Lithium batteries will offer the best growth opportunities in both the rechargeable and primary battery segments [44]. Figure 2.1 presents the Japanese battery market in the year 2013, in which lithium-metal technology is 17 % segment of the total battery revenue.

Table 2.4 Chemistry of various primary lithium batteries. Reproduced with permission from [45]. Copyright 1997 CRC Press

Fig. 2.1

Battery production in Japan for the year 2013. Primary batteries amounted to 61 %, including 17 % of lithium batteries

2.3.1 High Temperature Lithium Cells

2.3.1.1 Lithium Iron Disulfide Battery

The lithium iron sulfide battery operates at about 400–500 °C using a fused halide eutectic electrolyte immobilized in the pores of a suitable separator . This battery displays a number of attractive features compared to the sodium-sulfur battery, including prismatic flat-plate construction, ability to withstand numerous freeze-thaw cycles, cell failures in short-circuit conditions, ability to withstand overcharge and low-cost materials with available construction techniques. The major disadvantage is a somewhat lower performance. This battery is suitable for load-leveling applications. Attention has also focused on battery designs suitable for electric propulsion [46]. Energizer successfully produced the first commercially available AA-size 1.5 V Li-FeS₂ battery in 1989 (specific energy density ~297 Wh kg⁻¹). The 1.5 V AAA-size followed in 2005.

The most commonly used electrolytes are the LiCl-KCl binary eutectic and the ternary LiF-LiCl-LiI lithium halides. With Li-Al alloy anode and FeS₂ cathode the discharge occurs in several discrete steps:

$$ 4\mathrm{L}\mathrm{i}+3{\mathrm{Fe}\mathrm{S}}_2\to {\mathrm{Li}}_4{\mathrm{Fe}}_2{\mathrm{S}}_5+\mathrm{F}\mathrm{e}\mathrm{S}, $$

(2.2)

$$ 2\mathrm{L}\mathrm{i}+{\mathrm{Li}}_4{\mathrm{Fe}}_2{\mathrm{S}}_5+\mathrm{F}\mathrm{e}\mathrm{S}\to 3{\mathrm{Li}}_2{\mathrm{Fe}\mathrm{S}}_2, $$

(2.3)

$$ 6\mathrm{L}\mathrm{i}+3{\mathrm{Li}}_2{\mathrm{FeS}}_2\to 6{\mathrm{Li}}_2\mathrm{S}+3\mathrm{F}\mathrm{e}, $$

(2.4)

giving voltage plateaus at 2.1, 1.9, and 1.6 V, respectively. The use of Li-Al results in almost 50 % decrease in the theoretical specific energy, but much better stability is achieved. Most of the development was focussed on the LiAl/FeS couple; the Varta Battery Company (Germany) has produced a series of 140-Ah cells with a specific energy of 100 Wh kg⁻¹ at a low discharge rate of 80 mA cm⁻², falling to 50 Wh kg⁻¹ at high current density 250 mA cm⁻². Another version of LiAl/LiCl-LiBr-KBr/FeS₂ utilizes a dense FeS₂ electrode that only discharges to a stoichiometry corresponding to FeS (rather than to elemental Fe, which was the case of prior versions) [47]. The melting point of LiCl-LiBr-KBr at 310 °C allows cell operation at about 400 °C. These innovations have resulted in greatly improved capacity retention, and cells cycling for more than 1000 times. These cells are also developed in the USA and manufactured by Eagle-Picher, by Gould, and by SAFT America. Cells of 150- to 350-Ah capacity yield specific energies of 70–95 Wh kg^−l available at a 4-h discharge rate. There are still a number of unresolved scientific questions about the chemistry of LiAl/FeS cells and the mechanism of degradation and failure. In this system the separator is clearly a crucial component, which must not only keep the electrode materials apart but also allow good permeation of the electrolyte. The most suitable materials are found to be boron nitride and zirconia in the form of woven cloths, but they are obviously very expensive options.

2.3.1.2 Lithium Chloride Battery

Similar designs have been used in lithium chloride batteries, which operate at a temperature of 650 °C. The cell has the form Li(liq)/LiCl(liq)/Cl₂(g), carbon. The two electrodes, liquid Li anode and the porous carbon cathode in which the chlorine gas is fed under pressure, are separated by a molten lithium chloride electrolyte. The overall cell reaction is:

$$ \mathrm{L}\mathrm{i}\left(\mathrm{l}\mathrm{i}\mathrm{q}\right)+\mathrm{\frac{1}{2}}\kern0.5em {\mathrm{Cl}}_2\left(\mathrm{g}\right)\to \mathrm{Li}\mathrm{Cl}\left(\mathrm{l}\mathrm{i}\mathrm{q}\right), $$

(2.5)

with an OCV of 3.46 V. This system delivers a theoretical energy density 2.18 kWh kg⁻¹ at the working temperature. The most serious problems with this battery, however, are the corrosion of cell components and the development of satisfactory seals.

2.3.2 Solid-State Electrolyte Lithium Batteries

Solid-electrolyte batteries generally exhibit high thermal stability, low rates of self-discharge (shelf life of 5–10 years or better), ability to operate over a wide range of environmental conditions (temperature, pressure, and acceleration), and high energy densities of 0.3–0.7 Wh cm⁻³. However, limitations include relatively low power density due to the high impedance of most lithium-conducting solid electrolytes. The three commercial solid electrolyte battery systems are based on the solid electrolyte LiI, either formed in situ during cell manufacture or dispersed with alumina.

2.3.2.1 Lithium-Iodine Cells

From the first pacemaker implant in 1958 by Dr Ake Senning surgeon at the Karolinska Hospital in Stockholm, numerous engineering developments have faced challenges in battery power. In 1972, a primary lithium-iodine battery replacing the mercury-zinc cells greatly extended the cardiac pacemaker life (about 10 years). More details on the history of this battery can be found in ref. [48]. The most important factor for a cardiac pacemaker battery is its reliability [49]. The terminal voltage decay characteristic is well behaved, falling slowly enough for battery end-of-life (EOL) to be anticipated in routine follow up. The lithium-iodine battery has a solid anode of lithium and a polyphase cathode of poly-vinyl-pyridine (PVP), which is largely iodine (at 90 % by weight). The solid electrolyte is constituted by a thin film of LiI. The discharge reaction is given by:

$$ 2\mathrm{L}\mathrm{i}+n{\mathrm{I}}_2\left(\mathrm{P}\mathrm{V}\mathrm{P}\right)\to \left(n-1\right){\mathrm{I}}_2\left(\mathrm{P}\mathrm{V}\mathrm{P}\right)+2\mathrm{L}\mathrm{i}\mathrm{I}. $$

(2.6)

The cathode material is thus formed by the thermal reaction of iodine and polyvinyl-pyridine. This cell has an open-circuit voltage of 2.8 V. The theoretical specific energy for the Li/LiI/I₂(PVP) cell is 1.9 Wh cm^–3. The electrolyte ionic conductivity is ~6.5 × 10⁻⁷ S cm⁻¹ at 25 °C, and the energy density is 100–200 Wh kg⁻¹ [50]. The ratio of iodine to PVP is typically between 30/1 and 50/1, depending on the manufacturer. Phillips and Untereker have studied the phase diagram of the iodine/PVP material [51]. They demonstrated that at the application temperature (37 °C), the material passes through three phases—a two-phase region comprising a eutectic melt and pure iodine, a single phase liquid region, and a two-phase system wherein the iodine/monomer unit adduct phase coexists with the melt. An interesting and useful feature of the cathode material is that it is an electronic conductor. The conductivity is a function of the ratio of iodine to PVP. The maximum conductivity occurs at the I2/PVP weight ratio of 8/1. Since the initial ratio is much greater than this, the electronic conductivity increases until it reaches the weight ratio of 8/1, and then decreases gradually as the cell reaction continues. The increase in cell impedance caused by this decrease, together with the increase due to the formation of the lithium/iodine reaction product, leads to a gradual and predictable decrease in cell voltage that is easily detected by the electronic circuitry of the pacemaker. This feature allows the clinical personnel to detect the onset of the end-of-service point of the battery well before that point is reached, making it possible to schedule replacement surgery in a timely manner by routine checking of the battery status by telemetry. The volume change accompanying the cell discharge is ~12 % if the cathode is 91 % iodide by weight. This volume change may be accommodated by the formation of a porous discharge product or by the formation of macroscopic voids in the cell. Such batteries are used as power sources for implantable cardiac pacemakers, operating at 37 °C. They are commercialized by Catalyst Research Co., Wilson Greatbatch Inc., and Medtronic Inc. The lithium-iodine batteries have extended system lives up to 10 years for 120- to 250-mAh capacities. For use as power sources for portable monitoring or recording instruments, they have a nominal capacity of 15 Ah or less, and most have deliverable capacities under 5 Ah.

Batteries of medium capacities, i.e., up to around 1 Ah, can be used for random-access memory power supplies in electronics. Similar batteries using Li//Br have also been built. The greater electronegativity of bromine gives rise to voltages the order of 3.5 V and energy densities as high as 1.25 Wh cm⁻³. Their practical application is, however, limited by the low conductivity of the LiBr films that are formed.

2.3.2.2 Li/LiI-Al₂O₃/PbI₂ Cells

These batteries are recommended for low-rate operations and they are particularly suited for applications requiring long life under low drain or open-circuit conditions. Different cathodes have been used in these commercial solid-state cells, including a mixture of PbI₂ + Pb or PbI₂ + PbS + Pb, and a new system under development uses a mixture of TiS₂ + S or As₂S₃, which increases the energy density. The solid electrolyte is a dispersion of LiI and LiOH with alumina. Lithium ionic conductivities as high as 10⁻⁴ S cm⁻¹ have been reported in such a dispersion at temperature of 25 °C [52]. The discharge properties are characterized by an open-circuit voltage of 1.9 V and an energy density of 75–150 Wh kg⁻¹. A three-cell battery design manufactured by Duracell International delivers 6 V and offers a capacity of 140 mAh as a pacemaker power source. In typical complementary metal oxide semiconductor (CMOS) memory applications, the 350-mAh battery can be used with 1.0-V cutoff potential.

2.3.2.3 Carbon Tetramethyl-Ammonium Penta-Iodide Batteries

The Li/LiI (SiO₂, H₂O)/Me₄NI₅ + C cells were evaluated for use in cardiac pulse generators. They exhibited a voltage of 2.75 V and were projected to have a volumetric energy density of 0.4 Wh cm⁻³ [53]. The cathode is a mixture of carbon and tetramethyl-ammonium penta-iodide (Me₄NI₅).

2.3.2.4 Lithium Bromine Trifluoride Battery

Great advances have been made on the unconventional approach of combining alkali metals with strongly oxidizing liquid, e.g., SO₂, SOCl₂, or BrF₃, which acts simultaneously as electrolyte solvent and cathode depolarizer. BrF₃ is a very reactive liquid at room temperature, and the concept of a Li//BrF₃ cell has appeared [45, 53, 54]. A super high energy density battery which includes a lithium anode, a catalyst with liquid bromine trifluoride, and an electrolyte of antimony pentafluoride has been invented [55]. The cell reaction proceeds by the bromine trifluoride that disproportionates according the following equation:

$$ 4\mathrm{L}\mathrm{i}+{\mathrm{Br}\mathrm{F}}_3\to 3{\mathrm{Li}}^{+}{\mathrm{F}}^{-}+{\mathrm{Li}}^{+}{\mathrm{Br}}^{-}. $$

(2.7)

Consequently, no electrolyte salt is necessary because lithium was found to be stable in BrF₃ due to the formation of the protective surface layer. Upon cell activation a potential of 5 V is established, giving theoretical energy densities of 2680 Wh kg⁻¹ and 4480 Wh dm⁻³. A typical discharge of a Li/BrF₃/C cell is achieved at 5 mA cm⁻² current density, and a capacity of 5 mAh cm⁻² has been measured. The cell performance is controlled by the formation of the reaction product layer at the anode–electrolyte interface, leading to an increase in impedance. The BrF₃ electrolyte can also be modified by dissolution of various fluorides, e.g., LiAsF₆, LiPF₆, LiSbF₆, or LiBF₄. Such a cell is actually strongly developed by EIC Laboratories Inc. in the USA.

2.3.3 Liquid Cathode Lithium Batteries

2.3.3.1 Lithium Thionyl-Chloride Batteries

The lithium thionyl-chloride battery uses liquid thionyl chloride (SOCl₂) as its positive active material, and lithium metal as its negative active material [56]. The overall reaction of the battery is expressed as:

$$ 2{\mathrm{SO}\mathrm{Cl}}_2+4\mathrm{L}\mathrm{i}\to 4\mathrm{Li}\mathrm{Cl}+\mathrm{S}+{\mathrm{SO}}_2. $$

(2.8)

The Li//SOCl₂ battery achieves a high voltage of 3.6 V and provides high volumetric energy density of 970 Wh dm⁻³ with discharge current of 100 μA. It can be used over a wide temperature range from −55 to +85 °C. This cell displays remarkably lower self-discharge than conventional batteries owing to the formation of a LiCl protective layer over the Li-metal anode surface. Main applications of Li//SOCl₂ batteries are: medical instruments, cash registers, measuring instruments, onboard microcomputers, sensors, electronic meters (gas, water, electricity, etc. Bipower^® commercialized Li//SOCl₂ batteries with nominal voltage 3.6 V. The 1.6-Ah prismatic cell (BL–16PN-64) can be discharged at continuous current 20 mA. Applications include RF transmitters, military GPS systems, data loggers, alarm and security systems, CMOS memory backup and medical equipment. Li//SOCl₂ batteries from SAFT are bobbin (LS) or spiral (LSH) cells that operate at 3.6 V and show lowest self-discharge for extended operating life in the temperature range −60 to +150 °C. LS cells are designed specifically for long-term (5–20+ years) applications featuring a few μA base currents and periodic pulses, typically in the (−150 mA range, while. LSH cells are designed for applications requiring continuous current in the range 0.1–0.8 A. Taridan Batteries GmbH commercializes the SL-500 Li//SOCl₂ series exhibiting an extended temperature range up to +130 °C. These batteries have excellent shelf life (10 years) and extremely low self-discharge (1 % or less per year).

2.3.3.2 Lithium-Sulfur Dioxide Batteries

The cathode of the lithium-sulfur dioxide (Li-SO₂) batteries consists of a gas under pressure with another chemical as electrolyte salt; this is analogous to the thionyl chloride electrolyte and its liquid cathode [57]. Applications of these systems include emergency power units for aircraft and military cold-weather applications (e.g., radio operation). NASA is using Li-SO₂ cells for balloon and flight equipments. Features of Li-SO₂ batteries are as follows [58]: (1) high energy density up to 280 Wh kg⁻¹, (2) open-circuit voltage 2.95 V, operation voltage is between 2.7 and 2.9 V depending upon loading resistance, (3) long storage life, less than 2 % of self-discharging rate per year, could be stored for 10 years at room temperature, (4) spiral-wound construction may be discharged at higher rates, and (5) wide operation temperature range, −54 to +71 ° C. Spiral-wound Li-SO₂ battery from SAFT operates at 3 V in the temperature range −40 to +70 °C. This battery shows excellent capacity above 1 A. Its main advantage is the superior power at −40 °C and the low-self discharge during storage. The 11.5-Ah cell (model LO39SHX) can deliver 60 A maximum pulse discharge rate.

2.3.4 Solid Cathode Lithium Batteries

The semiconductive properties and tunnel structure of sulfide and transition-metal oxides led to the use of these materials in lithium power sources (Table 2.5). Several lithium-based chemistries were successfully applied to replace the prior system Zn/AgO and later the lithium-iodine batteries in implantable medical devices [59–61]. For example, Li//CuO, Li//V₂O₅, Li//CF _x , and more recently Li//Ag₂V₄O₁₁ couples have been adopted to power cardiac pacemakers requiring less that 200 μW [62, 63]. The lithium/carbon monofluoride (Li//CF _x ) primary cells are very attractive in several applications because of the double energy density with respect to the state-of-the-art Li//MnO₂ primary batteries (theoretically 2203 against 847 Wh kg⁻¹).

Table 2.5 Primary lithium batteries with solid cathodes

2.3.4.1 Lithium Polycarbon Fluoride Cells

Polycarbon fluorides of the general formula (CF _x ) _n can be obtained by direct fluorination of carbon black or other varieties of carbon at high temperatures. Theoretical specific energy density 2600 Wh kg⁻¹ can be achieved with these materials. Lithium cells with polycarbon fluoride cathodes have an open-circuit voltage in the range of 2.8–3.3 V, depending on the composition of the cathode material. A typical cell reaction may be written [64]:

$$ nx\mathrm{L}\mathrm{i}+\left(\mathrm{C}\mathrm{F},x\right)n\to n\mathrm{C}+nx\mathrm{L}\mathrm{i}\mathrm{F}. $$

(2.9)

Thus, the theoretical specific discharge capacity Q _th (in mAh g⁻¹) is expressed by:

$$ {Q}_{\mathrm{th}}(x)=\frac{xF}{3.6\left(12+19x\right)}. $$

(2.10)

The graph in Fig. 2.2 displays the experimental specific capacity of the (CF _x ) _n materials as a function of the stoichiometry (0.5 ≤ x ≤ 1.0). The theoretical capacity is calculated according to Eq. (2.11).

Fig. 2.2

Composition dependence of the experimental theoretical specific capacity of (CF _x ) _n cathode materials for primary lithium batteries. The theoretical capacity is calculated according to Eq. (2.11)

The crystal structure of graphite fluorides (CF _x ) _n compounds with x > 0.5 was investigated by Touhara et al. who proposed two phases: a stage-1 (CF₁) _n and a stage-2 (CF_0.5) _n , also commonly referred to as (C₂F) _n [65]. In stage-1 compounds, the fluorine is intercalated between each carbon layers to yield –CFCF- layers stacking, whereas in stage-2 (hexagonal, C _3h symmetry) according to their model, fluorine occupies every other layer with a stacking sequence of –CCFCCF–. The hexagonal symmetry is preserved in both (CF₁) _n and (CF_0.5) _n phases. Theoretical crystal structure calculations were also carried out and different layers stacking sequences were compared from their total energy. Due to electron localization in the C–F bond, a huge drop of the electrical conductivity occurs ~10⁻¹⁴ S cm⁻¹ in (CF) _n compared with ~1.7 × 10⁴ S cm⁻¹ in graphite.

Cells based on polycarbon fluorides are manufactured commercially in various forms. System developed by Matsushita Electric Industrial Co. is designed as a BR 435 cylindrical cell. Cells constructed by Nippon Steel Co. use carbon fibers as electrodes; they are found to be rechargeable. Cells for military applications have been produced in the USA by Eagle Picher and by Yardney Electric. There are spiral-wound cylindrical cells with a largest cell capacity of 5 Ah.

Lam and Yazami prepared sub-fluorinated graphite fluorides (CF _x ) _n compounds (0.33 < x < 0.63) from natural graphite. The cathode behavior in lithium batteries investigated under low discharge rate shows that the energy density increases with the fluorine content x. However, at higher rates, sub-fluorinated compounds performed better than commercial (CF _x ) _n synthesized from petroleum coke. Specific capacity 400 mAh g⁻¹ was obtained for Li//CF_0.52 coin cell discharged at 2.5C [66]. The rate capabilities of Li//(CF _x ) _n 18650-type 3.6-Ah cells using 1 mol L⁻¹ LiBF₄ in EC:PC:EMC (1:1:3) as electrolyte were tested at Sandia Natinal Labs in the temperature range −40 to +72 °C [67]. The Li//CF _x system was fabricated by Wilson Greatbatch as an alternative electrical source for advanced pacemaker systems [68]. Li//CF _x -MnO₂ hybrid cells are commercialized by Ultralife. These cells are fabricated for at least 15 years shelf life and deliver 40 % or more capacity than the same-sized Li//MnO₂ system. Li//CF _x is estimated to be ~9 % of the world’s primary battery market that is mainly shared by Panasonic (50 %), Greatbatch Medical (20 %), Spectrum Brands (20 %) and Eagle-Picher [69].

2.3.4.2 Lithium Manganese Oxide Batteries

Lithium manganese oxide (Li-MnO₂) battery is the most common consumer grade battery that covers about 80 % of the lithium battery market. This system includes heat-treated MnO₂ as cathode, lithium metal as anode and LiClO₄ in propylene carbonate/dimethoxyethane as aprotic electrolyte. The overall battery reaction is:

$$ {\mathrm{Mn}}^{\mathrm{IV}}{\mathrm{O}}_2+\mathrm{L}\mathrm{i}\to {\mathrm{Mn}}^{\mathrm{III}}\mathrm{OOLi}. $$

(2.11)

The cell reaction involves reduction of the MnO₂ without gases exhaustion and the product remains in the cathode. The system Li-MnO₂ displays several advantages: (1) the energy density is 150–250 Wh kg⁻¹ and 500–650 Wh dm⁻³, (2) the operating temperature ranges from −40 to +60 °C, (3) manganese dioxide is a low cost and safe material. Batteries are leading the market at approximately 50 % as estimated by Frost and Sullivan.

Many companies (Panasonic, Sony Energytec, Sanyo, Fuji, NTT, Matsushita, Varta, SAFT, etc.) embarked on ambitious programs to develop small rechargeable high-energy density, low cost Li-MnO₂ cells. Panasonic sales various battery sizes, from 30-mAh coin cells (CR1025) to cylindrical 2.4-Ah batteries. Duracell^® Li/MnO₂ batteries are being used in a wide range of applications, from powering all functions of fully automatic 35-mm flash cameras to providing long-term standby power for computer clock/calendars. The button cell and spiral-wound cylindrical cell Maxell^® Li//MnO₂ batteries have long-term reliability of 10 years; self-discharge rate is about 0.5 % per year. The cylindrical-type CR17450 model weighting 22 g has a nominal capacity 2500 mAh, V _oc = 3 V, and operates in the temperature range −40 to +85 °C. Typical voltage profiles of Li//MnO₂ batteries during discharge continuously at 40 mA are shown at different temperatures in Fig. 2.3 (from Maxell flyer). Scrap Li//MnO₂ batteries by Energizer^® meet US Department of Transportation (DOT) requirements of the 49CFR173.185 standard: limit of 12 g of Li per cell, strong outer packaging, external short circuits prevented. Sony has commercialized a button-type CR2025 cell tested at 23 °C under a load of 15 kΩ for 900 h. Excellent long-term storage reliability was demonstrated: at 60 °C for 120 days is the equivalent of storage at room temperature for 6 years. PowerStream fabricate CR series of primary Li//MnO₂ button cells with nominal capacity from 25 to 1000 mAh. The price of the CR2477 1 Ah-cell is $3.00. The cylindrical Li//MnO₂ 3-V battery by Bipower^® (model CR34615) of nominal capacity 8 Ah weights 120 g. Panasonic sells a varieties of light button cells; the CR2025 items of capacity 165 mAh weight 2.3 g. Since 1976, Sanyo Electric Co. has developed a wide variety of primary Li-MnO₂ cells to meet the demands of a large diversified market. Coin-type CR1220 36-mAh cell weighting 0.8 g delivers a maximum continuous discharge current 2 mA. The high-power cylindrical 2.6-Ah cell (spiral structure) weighting 23 g delivers 500 mA.

Fig. 2.3

Typical voltage profiles Li//MnO₂ batteries discharged continuously at 40 mA at different temperatures

2.3.4.3 Low Temperature Lithium Iron Batteries

Room temperature primary Li//FeS₂ batteries use nonaqueous electrolyte, such as a lithium/iron disulfide, with good low temperature performance characteristics. The electrolyte has a solute including lithium iodide dissolved in an ether-containing solvent including a 1,2-dimethoxypropane based solvent component and no more than 30 vol% 1,2-dimethoxyethane [70, 71]. Commercial Li//FeS₂ AA primary batteries have offered substantially longer runtime than equivalent sized Zn//MnO₂ alkaline AA cells for high-power applications and below ambient temperature [72]. The electrolyte consisted of 1 mol L⁻¹ LiCF₃SO₃ dissolved in 24.95 vol.% dioxolane, 74.85 vol.% dimethoxyethane, and 0.2 vol.% dimethylisoxazole. However, the rate capability of commercial Li//FeS₂ batteries may be limited by the ion conductivity of the electrolyte when FeS₂ samples have crystal sizes smaller than 1 μm. It has been demonstrated that the electrochemical performances are strongly dependent on the FeS₂ cathode particle size. Reducing crystal size of FeS₂ from 10 μm down to 100 nm increases the presence of crystal boundaries or surfaces per unit volume of FeS₂, provides a large number of nucleation sites for lithiated products, and thus promotes the kinetics of the heterogeneous (not intercalation) lithiation process of FeS₂ [73]. For a loading cathode of 14 mg cm⁻², the discharge voltage profiles of Li//nano-FeS₂ cells show a two-voltage-step reaction at 1.7 and 1.5 V at current densities lower than 3 mA g⁻¹ (0.042 mA cm⁻²). The 1.7-V lithiation plateau was attributed to the formation of pyrrhotite Fe_1−x S (0 < x < 0.2) and Li_2+x Fe_1−x S₂ (0 < x < 0.33) intermediate phases, and the 1.5 V reaction produced a mixture of plate-shaped Li₂S and amorphous Fe as final products.

2.3.4.4 Silver Vanadium Oxide Cells

Ag₂V₄O₁₁ (SVO) is an interesting cathode material for lithium primary batteries used as a power source for implantable cardiac defibrillators (ICDs). Li//SVO cells have traditionally been designed by Medtronic Inc. to discharge the cathode to a composition of >6Li:Ag₂V₄O₁₁ [74]. The SVO tunnel-like structure (C-centered monoclinic) consists of vanadium oxide layers with silver located between the layers. Li//SVO batteries have a distinctive discharge curve with two plateaus, one at 3.24 V and one at 2.6 V. In the fully discharged state, the composition of the cathode is Li₆Ag₂V₄O₁₁. This composition corresponds to reduction of silver to Ag⁰ and reduction of V⁵⁺ to V⁴⁺ [75, 76]. Lithium batteries with hybrid cathodes of Ag₂V₄O₁₁ and CF _x have been developed by Medtronic Inc. that combine the best features of both cathode components. Figure 2.4 compares the discharge profile of Li//SVO and Li//SVO-CF _x composite cells. Silver chromate, Ag₂CrO₄ was also used as a cathode material in Li primary battery for pacemaker. The cell reaction is identical to that of SVO with the formation of metallic silver and final solid product Li₂CrO₄. The nominal voltage is 3.5 V and the specific energy is 200 or 575 Wh dm⁻³ for a 2.5-V cutoff. Button cells were produced by SAFT.

Fig. 2.4

The discharge profile of Li//SVO and Li//SVO-CF _x composite cells used to power medical devices

2.3.4.5 Other Primary Lithium Batteries

In many cases the discharge mechanisms involved in lithium oxide cells are still not fully understood. The discharge reaction can be described as a formal displacement process:

$$ 2\mathrm{L}\mathrm{i}+M\mathrm{O}\to {\mathrm{Li}}_2\mathrm{O}+M, $$

(2.12)

where MO is an oxide such as CuO, Mn₂O₃, Bi₂O₃, or Pb₃O₄ with theoretical capacity of 670, 310, 350, and 310 mAh g⁻¹, respectively. OCV are in the range 3.0–3.5 V and a practical energy density of 500 Wh dm⁻³ is obtained. Varta has developed a CR 2025-type cell and SAFT has manufactured an LM 2020-type cell. The Li//CuO button cell (LC01) constructed by SAFT exhibits a single step which may be attributed to the simple displacement reaction (Eq. 2.12). This Li//CuO cell delivers an OCV of 1.5 V and has the highest specific energy of all solid cathode-based lithium cells. Practical value of 750 Wh dm⁻³ is obtained. The liquid electrolyte varies from manufacturer to manufacturer, but LiClO₄ in dioxolane is very often used. The cylindrical cells manufactured by SAFT have practical capacities in the range 0.5–3.9 Ah.

Sulfide electrodes have the advantage over the corresponding oxides that most of them are good electronic conductors, so that sulfide-based batteries usually do not require the addition of carbon into the cathode. Batteries based on cupric sulfide cells (three in series) have been developed for use in a cardiac pacemaker. Li//CuS undergoes in two steps, so that the discharge curve exhibits two plateaus at 2.12 and 1.75 V during the first discharge to 1.5 V with a capacity of 530 mAh g⁻¹. The reduction of CuS results in the formation of Li _x CuS during the first voltage plateau, and Cu_1.96S, Li₂S and metallic Cu during the second voltage plateau [77].

2.4 Secondary Lithium Batteries

Basically, the idea of using materials that undergo insertion reactions as the electrochemically active component at the positive electrodes began to be explored in the 1970s, giving birth to the rechargeable lithium batteries (RLB) . Two approaches consist in the design of RLB as shown in Fig. 2.4. The first system utilizes an insertion compound as positive material and a lithium-metal foil as the negative electrode, the so-called lithium-metal battery (Fig. 2.4a). The second system consists in using two open-structured materials as electrodes, in which the lithium ions can be shuttled from one intercalation compound acting as lithium-ion source to another receiving lithium-ion and vice versa for the discharge process. This is the commonly known lithium-ion battery (Fig. 2.4b) that was initially named “rocking chair battery,” “swing battery” or “shuttlecock battery.” Sometimes, LiBs are also named “lithium metal-free rechargeable batteries” [10] (Fig. 2.5).

Fig. 2.5

Schematic representation of rechargeable lithium batteries. There are two systems according the nature of the negative electrode either Li metal (a) or Li insertion compound (b). In both cases, the positive electrode is constituted by an insertion compound, in which the redox reaction occurs at high potential versus Li⁰/Li⁺

2.4.1 Lithium-Metal Batteries

The first electrochemical applications of intercalation compounds appeared on lithium-metal batteries (LMBs). The idea of using materials that undergo insertion reactions as the electrochemically active components of batteries began to be explored and accepted in the early 1970s [8]. Of particular interest was the case of an electron donor mechanism that takes place in transition-metal dichalcogenides MX ₂ (X = S, Se). The prototype is the Li//Li _x TiS₂ system [24]. Then, a series of studies of the transition metals of the first and second row were reported. Basically, batteries using MX ₂ compounds can be considered as concentration cells, in which the activity of lithium varies in the composition range 0 ≤ x ≤ 1. For x = 0, TiS₂ is at the pristine state and the cell is fully charged, while, for x = 1, the fully intercalated LiTiS₂ state corresponds to the discharged state. In Li _x TiS₂, for instance, the inherent reversibility of intercalation mechanism suggests the following reactions

$$ x{\mathrm{Li}}^{+}+x{e}^{-}+{\mathrm{TiS}}_2\rightleftarrows {\mathrm{Li}}_x{\mathrm{TiS}}_2. $$

(2.13)

As the intercalation of lithium occurs as ionic species Li⁺, the electro-neutrality of the host is maintained by the ejection into the external circuit of an electron generated by a reduction reaction of the transition-metal cation, the oxidation state of which switches from Ti⁴⁺ to Ti³⁺. The cell potential decreases upon the Li uptake into the host framework, which follows the simple Nernst equation:

$$ V(x)={E}^0-\frac{RT}{F} \ln a\left(L{i}^{+}\right), $$

(2.14)

where a(Li⁺) is the activity of Li⁺ ions in Li _x TiS₂. Note that the decrease of the cell potential with increasing of Li⁺ ion activity is a typical feature of Li-intercalation batteries. In the limits of the structural stability of the host, the electrochemical reaction is fully reversible.

Whittingham reported that the reversibility of the Li//TiS₂ couple permitted deep cycling for close to 1000 cycles with minimal capacity loss, less than 0.05 % per cycle, with excess lithium anode [78]. Two noticeable batteries with LiAl anode [79] that improves the safety were constructed by Exxon: the LiAl//TiS₂ cells were commercialized to power small devices such as watches and the largest lithium-metal prismatic-type cell using a LiB(CH₃)₄ salt in dioxolane was exhibited at the Electric Vehicle Show in Chicago en 1977 [78]. Because TiS₂ is a degenerate semiconductor that exhibits a high electronic conductivity, the use of this compound as cathode does not require the fabrication of composite electrode including conductive additive like carbon black. Despite the excellent performance of TiS₂, the great efforts of the Exxon’group were in vain, due to the formation of dendrites at the metallic anode in liquid electrolytes upon the cycling process, which is related to the non uniform change in the morphology of the anode surface as shown in Fig. 2.7. As pointed out by Xu [80], serious safety hazards are often caused by the generation of both dendrites and isolated lithium crystals (Fig. 2.6). The former creates internal shorts, and the latter is chemically active with the electrolyte solvents due to the huge surface areas of these lithium crystals. For the first discharge, that is the formation process, a passivation layer is formed at the surface of the lithium anode; this the solid electrolyte interphase (SEI), which is a poor ionic conductor in nature (Fig. 2.6a). This SEI is also growing upon repeated cycles. After the departure from the anode surface during discharge, the Li particles migrate to the cathode; the phenomenon is reverse for the charge. The deposition of Li⁺ ions coming from the cathode participates to the nucleation of dendrites, because particles do not have memory to seat back on the same site as before (Fig. 2.6b). Finally, the continuous dissolution and dendrite formation occur at the expense of the mechanical stability of the anode, which damages the whole battery operation.

Fig. 2.6

Sketch of the formation of dendrites and deposition of lithium grains at the surface of the electrode (a) for the first discharge process with formation of the SEI and (b) upon cycling

The use of LMBs remains inefficient because several factors that are: (1) the unresolved control of the reactivity of lithium toward electrolyte, (2) the difficult mastery of the SEI layer, (3) the formation of large dendrites causes internal short circuiting, and (4) the increasing surface area upon cycling poses risks of overheating. It is a fact that LMBs have low coulombic efficiency ~90 % with carbonate solvents, and low cycle life due to the ceaseless growth of the passivation layer [81]. Von Sacken et al. [82] reported the thermal runaway of anode materials and showed that reactions between lithium metal and electrolyte are exothermic. These aspects are developed in the following, in the chapter related to anode materials.

Beside TiS₂, several transition-metal chalcogenide MX ₂ or MX ₃ (X = S, Se) compounds with a layered structure (case of MX ₂) of one-dimensional structure (NbSe₃ has been the model system for quasi-1D charge density waves) have been studied as cathode materials in LMBs. Most of them display a single-phase behavior upon lithium intercalation. Li _x VSe₂ is the unique material showing a two-phase behavior that is evidenced by the appearance of two voltage plateaus in the discharge curve [83]. Three systems had emerged using TiS₂, MoS₂ and NbSe₃. A solid state battery Li/chalcogenide glass/TiS₂-carbon was developed at Eveready Battery Co. (USA) with the target of CMOS memory backup market. The electrolyte is a phosphorous chalcogenide glass Li₈P4_0.25S_13.75 mixed with LiI and on a composite electrode of TiS₂-solid electrolyte-black carbon with percentages in weight 51:42:7. The capacity ranges from 1.0 to 9.5 mAh. The cell packaging is a standard-sized XR2016 coin cell. The resistance of the cell at 21 °C is between 25 and 100 Ω depending on the state of charge [84]. A spirally wound AA-size Li/TiS₂ cell was constructed by Grace Co. (USA). At 200-mA discharge rate, a capacity of l Ah is delivered to 1.7 V [33]. Capacity of 1.6 Ah is obtained from a C-size Li/TiS₂ cell built by EIC Laboratories Inc. (USA). This cell operates in the temperature range from −20 to +20 °C [34]. C-size 3.7-Ah Li/MoS₂ batteries were especially manufactured by Moli Energy Ltd. (Canada) as the power source for pocket telephones in Japan. This battery uses a lithium anode, a propylene carbonate-based electrolyte solution and a processed MoS₂ cathode. Sustained drain rates of several amperes at a cell voltage between 2.3 and 1.3 V are obtained. The energy density is in the 60- to 65-Wh kg⁻¹ range at a discharge rate of C/3 (approximately 800 mA) [85]. The Li-NbSe₃ system, termed FARADAY cell [86] was developed at AT&T (USA). The AA-size cylindrical cell was designed for operating over 200 cycles at a typical current of 400 mA to a cutoff capacity of 0.7 Ah. The ability to incorporate 3Li per NbSe₃ gives a relatively high theoretical energy density of 1600 Wh dm⁻³. A practical energy density of 200 Wh dm⁻³ was achieved with the possibility of 350 cycles of charge–discharge.

It has long been recognized that lithium can be inserted chemically into metal oxide frameworks. Both molybdenum oxides [87, 88] and vanadium oxides [89] were synthesized for this purpose. The reversibility of insertion reaction in these alkali metal bronzes Li _x MO _y formed in electrochemical cells allows many of these cells to be rechargeable. Labat et al. [90] disclosed a rechargeable cell having a cathode based on V₂O₅. A C-type Li//V₂O₅ 1.4-Ah cell has been commercialized by SAFT. This cell includes a lithium or lithium alloy anode so that the cell is charged to 3.8 V. In the cell constructed by Tracor Applied Science Inc. about 2–20 % by weight of carbon is admixed to V₂O₅ [91]. The Li//V₂O₅ coin cell developed by Matsushita Micro Battery Corp. (Japan) provides power to electronic equipments like memory backup. The cathode is a mixture of V₂O₅ and 5 wt% carbon black. During the discharge, a depth of 3Li per V₂O₅ is obtained with an average voltage about 2 V. This battery delivers a capacity of 36 mA at a discharge rate of 1 mA [92].

2.4.2 Lithium-Ion Batteries

2.4.2.1 Principle

Basically, a lithium-ion battery (LiB) does not contain lithium metal, but only charge species Li⁺ ions. These ions swing from one electrode to another throughout the electrolyte that is a good ionic conductor and an electronic insulator. Let us consider the charge and discharge reaction for a LiB with the Li _x CoO₂/LiPF₆-EC-DMC/Li _x C₆ electrochemical chain as shown in Fig. 2.7. A fresh cell is in the discharge state, i.e., at low potential. Consequently, in the initial state, the positive electrode framework is full of Li⁺ ions (Li₁CoO₂), while the anode is empty (carbon C). As any electrochemical reaction implies the transport of ions and electrons, the redox process in a LiB is as follows. During the charge process, Li⁺ ions are generated by the positive electrode (anode at this time), migrate across the electrolyte and penetrate the negative electrode (cathode at this time), while electrons circulate through the external circuit; in the process, the positive electrode is oxidized losing x electrons (Li_1−x CoO₂) and the negative electrode is reduced capturing x electrons (Li _x C₆) and vice versa for the discharge process.

Fig. 2.7

Principle of the lithium-ion battery (a) charge process, (b) discharge process

For a LIB made of LiCoO₂ as positive electrode and graphite-like carbon as negative electrode, the chemical reactions for charge and discharge are expressed as shown below:

$$ x{\mathrm{Li}}^{+}+x{e}^{-}+6\mathrm{C}\rightleftarrows {\mathrm{Li}}_x{\mathrm{C}}_6\left(\mathrm{at}\ \mathrm{the}\ \mathrm{negative}\right), $$

(2.15)

$$ {\mathrm{Li}\mathrm{CoO}}_2\rightleftarrows {\mathrm{Li}}_{1-x}{\mathrm{CoO}}_2+x{\mathrm{Li}}^{+}+x{e}^{-}\left(\mathrm{at}\ \mathrm{the}\ \mathrm{positive}\right), $$

(2.16)

where the upper arrow represents the discharge process and the lower one the charge process. The overall battery reaction is expressed as:

$$ {\mathrm{Li}\mathrm{CoO}}_2+6\mathrm{C}\rightleftarrows {\mathrm{Li}}_{1-x}{\mathrm{C}\mathrm{oO}}_2+{\mathrm{Li}}_x{\mathrm{C}}_6. $$

(2.17)

Note that the LiCoO₂//C-type cell operates at high voltages, around 4 V, due to the big difference [μ(LiCoO₂) − μ(C)] between the chemical potentials of the electrodes (Eq. 1.2).

Currently, the electrolyte is a mixture of alkyl carbonate solvents (aprotic) with lithium hexafluophosphate salt (LiPF₆) to provide ionic conductivity for the back and forth motion of the lithium ions. Generally the transport of ions into a solid-state compound is a slow process that requires optimization by using very well crystallized electrode materials [93, 94]. However, some amorphous substances can be used and this area is receiving increasing attention, in particular in the case of some anode materials that are discussed in Chap. 10. An important strategy in the design and optimization of an electrode is the use of the nanotechnology to reduce the size of active particles as small as possible. The smaller the particle size, the shorter the length of the path of a guest species in the solid, and the smaller the change from the core to the surface of individual particles during discharge–charge cycling [95].

Lithium ion batteries have a high energy density and are ideal for cyclic applications. They allow savings in volume and weight up to 70 % compared to traditional lead acid batteries or Ni-MH batteries (see Table 2.6). One disadvantage of the Li-ion batteries is that Battery Management System (BMS) is inevitable, to control and monitor the battery pack, and automatically compensate imbalances between cells as much as possible. This ensures a constant high capacity and a long life span. The BMS must also test constantly each cell to avoid thermal runaway and the propagation to other cells (battery fires). However, the recent investigations of LiBs with iron phosphate and titanate electrodes have shown the LFP/LTO cell passes successfully the battery tests for safety use in public transportation [96], in which case the role of the BMS is reduced to the control of the balance between the cells in the battery pack. The “18650” battery prepared under such conditions delivers a capacity of 800 mAh. It retains full capacity after 20000 cycles performed at charge rate 10C (6 min), discharge rate 5C (12 min), and retains 95 % capacity after 30000 cycles at charge rate 15C (4 min) and discharge rate 5C, both at 100 % depth of discharge (DOD) and 100 % state of charge (SOC). Since the introduction of LiBs using LiCoO₂ as the oxidant electrode, several new cell chemistries have been developed and used in commercial batteries.

Table 2.6 Characteristics of Li-ion against NiMH cell

2.4.2.2 Energy Diagram

The open-circuit energy diagram of a lithium battery has been discussed by Goodenough and Kim [97]. Figure 2.8 represents the energetic configuration of electrodes and electrolyte of a lithium battery at the thermodynamic equilibrium. The anode and cathode electrodes are electronic conductors with electrochemical potentials μ _A and μ _C, also called Fermi level energies E _A and E _C, respectively. The electrolyte is an ionic conductor with a band gap E _g, which represents the separation between the lowest unoccupied molecular orbital (LUMO) of energy E _L and the highest occupied molecular orbital (HOMO) of energy E _H. The thermodynamic stability of a lithium cell requires that the electrochemical potentials of electrodes E _A and E _C are located within the energetic window of the electrolyte, which constrains the cell voltage V _oc of the electrochemical cell to

Fig. 2.8

Electronic band diagram of Li-ion batteries with various electrodes: olivine LiFePO₄, lamellar LiCoO₂, and spinel LiNi_1/2Mn_3/2O₄ as cathodes and lithium-metal and carbon as anodes. E _A and E _C represent the Fermi level of anode and cathode respectively. E _g is the electrolytic window that ensures the thermodynamic stability, while E _A > E _L and E _C < E _H requires a kinetic stability by the formation of an SEI layer

$$ e{V}_{\mathrm{oc}}={E}_{\mathrm{C}}-{E}_{\mathrm{A}}\ge {E}_{\mathrm{g}}, $$

(2.18)

where e is the elementary electronic charge and E _g = E _L − E _H [14].

It has been suggested that in practical nonaqueous lithium battery systems the anode (Li or graphite) is always covered by a surface layer named the solid electrolyte interphase (SEI), 1–3 nm thick, which is instantly formed by the reaction of the metal with the electrolyte (this is called “formation cycle”). This film, which acts as an interphase between the metal and the solution, has the properties of a solid electrolyte. The SEI acts as a passivating layer at the electrode/electrolyte boundary gives a kinetic stability to the cell only when the condition in Eq. (2.18) is fulfilled. Otherwise, this layer has a corrosive effect and grows with the cycling life of the battery [98]. The design of electrodes must match the LUMO and HOMO level of the electrolyte. In Fig. 2.8, we present the schematic energy diagram of Li-ion cells for three different chemistries that illustrate the different situations: titanate spinel-iron phosphate (LTO//LFP), graphite-lithium cobaltate (C//LCO) and graphite-nickel-manganese spinel (C//LNM). For the couple Li₄Ti₅O₁₂//LiFePO₄, the voltage cell V _oc is smaller than E _g. Therefore, no SEI is formed because the electrode energy levels E _C and E _A match well within the electrolytic window, which insures very high safety. However, the price to be paid is a lower open-circuit voltage, i.e., 2 V against 4 V for the graphite//LiCoO₂ battery. For the graphite//LiCoO₂ cell, graphite has E _A lying above the LUMO of used nonaqueous electrolytes and E _C of the cathode lies below the HOMO level. Consequently, the use of both graphite and LiCoO₂ electrodes is possible because this combination allows for the growth of the passivating SEI film. The SEI requires properties as follows: (1) it must have good mechanical stability when changes in electrode volume occur upon cycling life, (2) it must allow for fast Li⁺-ion transfer from the electrolyte to the electrode, and (3) it must have a good ionic conductivity over the temperature range −40 < T < 60 °C [99]. In the case of the graphite//LiNi_1/2Mn_3/2O₄ cell, the situation is worse because E _C lies far from E _H that makes the electrode-electrolyte very instable. Table 2.7 lists a selection of the most popular Li-ion technologies that have been developed so far.

Table 2.7 The most popular Li-ion technologies developed so far

The characteristics of the cathode and anode elements together with a discussion on their advantages and disadvantages are reported in following chapters of this book. In Table 2.7, however, only the Li₄Ti₅O₁₂ (LTO) material is reported as an alternative to graphite, since it is being commercialized. LTO has been considered as a viable 1.5 V anode material since some time due to several advantages (see Chap. 10), and LiBs with this anode and various cathodes have been tested, starting with the LTO//LiMn₂O₄ combination, with liquid [100] or solid polymer [101] electrolytes. Then, the combinations LTO//LiNi_0.8Co_0.2O₂ and LTO//LiCoO₂ have been explored [102], demonstrating their viability for high-power applications. However, the exploration of LTO-based batteries was more or less abandoned for some years, due to the focus of all the research on only one parameter: the energy density and LTO suffered from the lower operating voltage. The interest in LTO was revived by the Ohzuku group [103, 104] who showed that a combination of LTO with a Li mixed oxide as cathode can deliver gravimetric energy densities as high as 250 Wh kg⁻¹ and volumetric energy densities of 970 Wh dm⁻³. The LTO//LiMn₂O₄ cell delivers a reversible capacity 90–100 mAh g⁻¹ when cycled at 55 °C at 1C rate [105, 106]. With LTO particles of diameter 20 nm, the nano-LTO//LiMn₂O₄ cell can be cycled up to 1000 cycles at 25 and 55 °C and fast-charge capability was reported [107] up to 80C rate. The best results for the LTO//LiMn₂O₄ cell have been obtained with micrometer sized (~0.5–2 μm) secondary particles composed of nanosized (≤10 nm) LTO primary particles as the anode, combined with spherical LiMn₂O₄ particles [108]. This cell showed nearly 100 % capacity retention up to 1000 cycles when cycled at 5C rate at 55 °C. However, the problem with the use of LiMn₂O₄ is not the cycling life (number of cycles before aging), but the calendar life, which is very limited by the dissolution of manganese in the electrolyte. In addition, as any chemical process, the kinetics of this dissolution process increases with temperature. An illustration of this effect was given by the failure of LiMn₂O₄-based batteries that equipped electric cars in the US after a hot summer in 2013, which forced the car-maker to recall these cars to change the batteries. That is why many efforts were made since 10 years to replace LiMn₂O₄ by a compound based on another transition metal that does not dissolve into the electrolyte. The LTO//LiFePO₄ (LFP) cell operates at voltage circa 1.75 V and has been first studied in 1994, using a gel-polymer electrolyte [109]. A liquid electrolyte made of Li salt in an ionic liquid, namely, lithiumbis(trifluoromethane) sulfonamide, LiTFSI in Py₂₄TFSI, is also viable [110]. When the LiFePO₄ particles used as cathode of the LTO//LFP cell are coated with a conductive layer to increase the electrical conductivity of the LFP powder, the cell delivers a capacity of 155 mAh g⁻¹. In one case, the coat was polyacene [111], but usually it is simply a 2–3 nm thick conductive carbon layer [112]. A breakthrough performance has been obtained with such carbon-coated nano-LFP (particle size, 25 nm) as cathode [96]. The cells retained full capacity after 20000 cycles performed at charge rate of 10C and discharge rate of 5C, and retained 95 % of the capacity after 30000 cycles at charge rate 15C and discharge rate 5C, both at 100 % DOD (depth of discharge) and 100 % SOC (state of charge). The performance has even been improved by using carbon-coated LTO particles of size 90 nm, as good rate capability has been observed up to 100C rate [113]. In addition, different electrolytes have been tested in this work in order to increase the temperature where the battery can operate, which imposes to get rid of the ethylene carbonate that has a too small boiling temperature. The upper bound of the temperature window where the “18650” battery operates is raised to 80 °C by using 1 mol L⁻¹ lithium bis(fluorosulfonyl)imide (LiFSI) in GBL or 1 mol L⁻¹ LiFSI in PC+ γ-butyrolactone (GBL) electrolyte. Studies have also been successful to decrease the lower limit of the temperature window. The “18650” cell using the low temperature electrolyte 1 mol L⁻¹ LiPF₆ and 0.2 mol L⁻¹ LiFSI in quaternary blend of aliphatic solvents such as propylene (PC) + methyl propionate (MP) + ethylmethyl carbonate (EMC) + 5 % fluorinated ethylene carbonate (FEC) (1:1:1:1 by volume) displayed a similar electrochemical performance at 25 °C as that of the cell with conventional electrolyte, but in contrast permits to pass the Hybrid pulse power characterization (HPPC) test down to −10 °C, owing to improvement in ionic conductivity of the electrolyte. In addition, accelerated rate calorimetry measurements have shown that this cell benefits from an unprecedented safety [113]. The LTO/LFP combination then appears as the most attractive battery for use in PHEV, and energy storage for wind and solar energy. As mentioned earlier, the LiNi_1/2Mn_3/2O₄ (LNM) cathode, which operates at 4.7 V with respect to Li, cannot be combined with graphite, and LTO has then been considered as an alternative anode by the Ohzuku group [104, 114, 115]. The LNM//LTO cell operates at 3.2 V and delivers a capacity decreasing from 5.4− 4.8 mAh (for mass balanced cells of 0.056 g LTO and 0.049 g LNM) when the current rate increases from 12.7 to 509 mA g⁻¹. The Amine group found a 86 % capacity retention when the LNM//LTO cell was cycled at C-rates from 0.5 to 10C, with very good cyclability over 1000 cycles between 2.0 and 3.5 V [116]. A Li(Ni_0.45Co_0.1Mn_1.45)O₄//LTO with weight ratio of anode to cathode 1.36 in the laminated-type cell delivered a capacity of 124 mAh g ¹ up to 500 cycles at 1C rate [117]. Let us recall, however, that the problem with LNM (Co-doped or not) is the dissolution of manganese in the conventional electrolytes (same problem as in the case of the LiMn₂O₄) that reduces the calendar life and is still an obstacle to the commercialization of such batteries.

LTO has also been combined with the layered oxides LiCoO₂ [102], Li(Ni_0.8Co_0.2)O₂, Li(Co_1/2Ni_1/2)O₂ [118], Li(Co_1/3Ni_1/3Mn_1/3)O₂ [104], and Li_1+x [Co_1/3Ni_1/3Mn_1/3]_1−x O₂ [119]. The performance of such batteries is good. The interest in the use of such layered compounds as cathodes is the high energy density. For instance, LTO//Li(Co_1/3Ni_1/3Mn_1/3)O₂ cell delivered a capacity <85 mAh g⁻¹ (340 mAh cm⁻³) and has an energy density of 215 Wh kg⁻¹ or 970 Wh dm⁻³ with an average voltage of 2.5 V. The replacement of graphite by LTO improves the power of the battery, since this material does not change volume upon insertion of lithium, and supports much faster charge–discharge rates than graphite. The safety problem with layer compounds, however, remains. Their bad thermal stability has restricted the use of these materials as cathode of LiBs to applications to portable use, with the exception of Boeing and Tesla, who used Al-doped Li(Ni_0.8Co_0.2)O₂//graphite, inevitably leading to the battery fires that they have experienced if the BMS cannot manage the intrinsic instability of these batteries. That is why we did not recommend their use in transportation in spite of the temptation of the higher energy density.

Other batteries have been studied at the laboratory scale, using other compounds that are extensively reviewed in Chap. 10 as promising anode elements. They are, however, still too far from commercial use to be listed in Table 2.7. In addition, as we see in Chap. 10, the performance of these materials depend very much on the composition of the anode electrodes, and the research today is focused on their optimization by working on the formation of composites associating for instance carbon under the form of nanotubes or graphene, or coat of the nanoparticles of active elements of different shape (spherical, nanotubes, nanoplates), and different porosity. A review of primary results on such full cells (by opposition to half-cells with Li as the anode studied in Chap. 10) can be found for instance in ref. [120], but they will be fast outdated by the progress on the building of the anodes and the results newly obtained on half-cells discussed in Chap. 10, owing to the progress in nanoscience and materials science.

2.4.2.3 Design and Manufacturing

The elementary electrochemical cell of the lithium-ion battery is based on the assembly of three main components. This cell comprises two reversible electrodes that are slurry deposited onto metallic current collector: the anode provides lithium ions during discharge and the cathode receives these charge species which are inserted into its lattice. The two electrodes are separated by a porous membrane with foam structure (separator) [121] soaked with the liquid electrolyte. This electrolyte is a fast ion conductor of lithium ions, but it is also an electronic insulator, because any transport of electron across the electrolyte would result in a self-discharge. The ionic conductivity is achieved by dissolving lithium salts in aprotic solvents. The anode is deposited onto copper foil, while the cathode material that is a mixture of electro-active material, carbon black and binder, is coated onto an aluminum foil. Both metallic foils insure the function of current collector. Each electrode film spread by coating technique has a thickness of a few tens of microns and a width of a few centimeters. Numerous operations are made to construct a cell, from preparation of powders and coating to cell formation and electrical tests (Fig. 2.9). The total thickness of the laminate obtained is about 150 μm, which depends on the electrical characteristics of the required cell. Indeed, changing the film thickness maximizes performance to meet the specific request of the application. So assemble thick film provides a configuration of high-energy type battery, while a thinner element will provide high-power performance. For a brief overview of processing for lithium-ion batteries see Ref. [122]. The sensitive chemistry of the Li-ion cell means that the manufacture must be done in dry room, less than 100 ppm of water. Mass production lies in fully automatic lines.

Fig. 2.9

Chart of manufacturing a Li-ion battery

The manufacture of single cells and modules and packs for large-format power batteries is under development. One thing is inescapable: the manufacturing of cells in a dry room due to the sensitivity of lithium and chemistry to moisture. To satisfy the requirement of the target market of batteries for electric vehicles, the fabrication should be fast and less expensive, which necessitates fully automatic and integrated production lines. Such a solution has been adopted at Hydro-Québec, Canada, manufacturing 10-Ah iron-phosphate lithium-ion cells that are assembled in pack for power applications such as HEV, EVs and stationary storage. Figure 2.10 compares the performance of various battery technologies envisaged for HEV and EV applications. Except energy and power density markers, the LFP chemistry shows the best characteristics in terms of fast charge, cycling, temperature runway and safety.

Fig. 2.10

Challenges in electrode materials for Li-ion batteries. LFP lithium iron phosphates. NMC nickel-manganese-cobalt oxide. NCA nickel-cobalt-aluminum oxide. LMS lithium-manganese spinel. Ranking: 1 = worst, 5 = best

2.4.3 Lithium Polymer Batteries

In the 1980s Armand in collaboration with ELF-Aquitaine (France) and Hydro-Québec (Canada) played a leading role in the development of the lithium polymer battery (LMP) technology. Armand [123] suggested that solid polymer electrolytes derived from polyethylene oxide (PEO) might be used as electrolytes in LMP batteries, and major efforts have been made to produce polymer electrolytes with high conductivities at room temperature. Polymeric electrolytes are formed by complexes between salts of alkali metals and polymers containing solvating heteroatoms localized on the polymer chains. The most common example concerns such complexes with PEO chains that are formed by the repeat unit –CH₂–CH₂–O–. The solvating heteroatom, here oxygen, acts as a donor for the cation M ⁺. The anion X ⁻, generally of large dimension, stabilizes the PEO-M ⁺ complex. The first polymer electrolytes had conductivities of less than 10⁻⁵ S cm⁻¹ at 25 °C, far too small for use in normal battery applications. Recently, however, electrolyte compositions have been produced which exhibit conductivity characteristics competitive with the properties of nonaqueous liquid electrolytes [124]. These materials have stimulated major development efforts in polymer electrolyte battery technology. The PEO-MX complexes may be considered as plastic electrolytes that are a compromise between liquid and solid crystalline electrolytes. Most of the polymeric batteries are of the form Li/PEO-Li salt/IC, where IC can be an intercalation compound, but also a composite electrode including a large amount of active intercalation material. From an examination of the temperature dependence of the ionic conductivity in polymer complexes, one can remark the relatively high conductivity at room temperature of the complexes based on polyphosphazene, due to the low value of the vitreous transition temperature, T _g, and thus to the flexibility of the chains of the polymer, in turn related to the small barrier height to rotation of the P-N bond [125].

Positive electrodes of LPM batteries include a lithium-insertion electrode such as V₆O₁₃, TiS₂, MnO₂, Cr₃O₈, or LiV₃O₈ admixed with polymer electrolyte and conducting carbon for enhancement of the ionic and electronic conductivities, respectively. The composite electrode (50- to 75-μm thick) is deposited on a thin copper or nickel foil as current collector of few μm thick, and a film (25–50 μm) of [(C₂H₄O)₉⋅LiCF₃SO₃] _n polymer electrolyte completes the cell (Fig. 2.11). For optimum conductivity, the temperature of the polymer is maintained in the temperature range 80–90 °C. The system examined in the French–Canadian Project considered the following sequence Li/(PEO)₈⋅LiClO₄/TiS₂ + PEO + C [124]. Another system including TiO₂ as the cathode was cycled in the voltage range 3.0–1.2 V at a C/8 rate. A comparison of both cells shows similar features.

Fig. 2.11

Structure of a prismatic lithium-ion battery

Recently, a LMP battery pack has been developed by Batscap that aims to demonstrate the feasibility of the high power LMP technology for EV application. It was especially designed to meet the needs for electric vehicles. This module, consisting of 12 cells connected in series is equipped with an electronic system (BMS) that provides thermal management (control of the internal temperature) and the electrical operation. The main information relating to security (alarm management) and the state of charge are managed and communicated to the application. With features in specific and volumetric energy density above 100 Wh kg⁻¹ and 100 Wh L⁻¹, respectively, the 2.8-kWh BMP module (rated voltage 31 V) developed by BatScap realizes outstanding performance of lightness (25 kg) and compactness (25 L) and successfully equipped the fully electric Bluecar of the Bollore group.

2.4.4 Lithium-Sulfur Batteries

The lithium-sulfur (Li-S) battery has been under intense scrutiny for over two decades, as it offers the possibility of high gravimetric capacities and theoretical energy densities. Sulfur has a high specific capacity of 1673 mAh g⁻¹, but the rapid capacity fading due to dissolution of polysulfides poses a significant challenge for practical applications. The discharge process of Li-S cell occurs with lithium dissolution from the anode surface and insertion into alkali-metal polysulfide salts, and reverse lithium plating to the anode while charging [126]. Theoretically, each sulfur atom can host two lithium ions, but experiments show that only 0.5–0.7 Li⁺ ions are accommodated. Consequently, polysulfides are reduced while the cell is discharged according the reactions [126–129]:

$$ {\mathrm{S}}_8\to {\mathrm{Li}}_2{\mathrm{S}}_8\to {\mathrm{Li}}_2{\mathrm{S}}_6\to {\mathrm{Li}}_2{\mathrm{S}}_4\to {\mathrm{Li}}_2{\mathrm{S}}_3, $$

(2.19)

across a porous diffusion separator, sulfur polymers form at the cathode as the cell charges:

$$ {\mathrm{Li}}_2\mathrm{S}\to {\mathrm{Li}}_2{\mathrm{S}}_2\to {\mathrm{Li}}_2{\mathrm{S}}_3\to {\mathrm{Li}}_2{\mathrm{S}}_4\to {\mathrm{Li}}_2{\mathrm{S}}_6\to {\mathrm{Li}}_2{\mathrm{S}}_8\to {\mathrm{S}}_8. $$

(2.20)

These reactions are analogous to those in the Na/S battery. Octasulfur has three forms: α-sulfur, β-sulfur, and γ-sulfur; the β-sulfur and γ-sulfur are metastable as they convert to α-sulfur in storage at ambient temperature. A typical discharge and charge voltage profile of the first cycle of Li//S cells is shown in Fig. 2.12.

Fig. 2.12

Typical discharge and charge voltage profile of the first cycle of Li//S cells

Different configurations of Li//S batteries include silicon or lithium anode and a varieties of sulfur-based cathodes such as pure sulfur, porous TiO₂-encapsulated sulfur nanoparticles, sulfur-coated, disordered carbon nanotubes made from carbohydrates, copolymerized sulfur, sulfur-graphene oxide nanocomposite [129–132]. A 180-Å thick carbon layer was coated onto sulfur cathodes prepared by sputtering method. Charge–discharge tests show a specific capacity 1178 mAh g⁻¹ at first discharge, falling to about 500 mAh g⁻¹ after 50 cycles [132]. Ji et al. [130] reported the feasibility of a highly ordered nanostructured carbon-sulfur cathode for lithium-sulfur batteries. The conductive mesoporous carbon framework precisely constrains sulfur nanofiller growth within its channels and generates essential electrical contact to the insulating sulfur. Such a structure favors access to Li reactivity with the sulfur. Reversible capacities up to 1320 mAh g⁻¹ were attained. Zheng et al. [131] reported a hollow carbon nanofiber-encapsulated sulfur cathode for effective trapping of polysulfides, fabricated by using anodic aluminum oxide templates, and thermal carbonization of polystyrene. A high specific capacity of about 730 mAh g⁻¹ was observed at C/5 rate after 150 cycles of charge–discharge. The introduction of LiNO₃ additive to the electrolyte was shown to improve the coulombic efficiency to over 99 % at rate C/5.

2.5 Economy of Lithium Batteries

Let us consider several aspects of lithium battery managements. Figure 2.13 displays some battery sizes including cylindrical, coin-cell, prismatic and pouch cell. A 2-Ah 18650 Li-ion cell has 0.6 g of lithium content. On a typical 60 Wh laptop battery with 8 cells (4 in series and 2 in parallel), this adds up to 4.8 g. To stay under the 8-g UN limit, the largest battery you can bring is 96 Wh. This pack could include 2.2 Ah cells in a 12 cells arrangement (4 series-3 parallels). If the 2.4-Ah cells were used instead, the pack would need to be limited to 9 cells (3 series-3 parallels). Evidently, the eventual commercial acceptability of lithium rechargeable batteries will depend on their cost effectiveness compared with other types of battery in the same utilization. In France, the cost of domestic electricity is 1.31€ per kWh including VAT and taxes (in January 2014); a AA-size alkaline Leclanché cell delivering, say 5 Wh, currently retails at 1.50€. Thus energy from this primary battery costs 300€/kWh, i.e., more expensive by a factor of over 200 with respect to the energy provided by domestic electricity.

Fig. 2.13

Various battery configurations. (a) LS14250- cylindrical Li//SOCl₂ battery, (b) 4R25-prismatic (container) cell, (c–d) CR2020 Li-MnO₂ coin cells, (e) AA-type Li-ion cylindrical (1.5 Ah, 3.6 V, 12.4 g) cell, and (f) 30-Ah pouch (LiMn₂O₄-type, 3.8 V, 850 g) cell

Information from battery manufacturers are rather sparse but prices have now slipped to anywhere from $400 to $750 a kWh that means the 85-kWh pack powering the Tesla electric car costs between $34000 and $6375. A study by the Boston Consulting Group projected that prices would need to come down to $200 or less per kWh to make electric vehicles truly competitive. With a forecast of global demand of 26149 MWh by 2015, Sam Jaffe, research manager at IDC Corporate USA, claims a dramatic reduction in price for Li-ion cells to as low as $400/kWh [133]. Every year over 15 billion electrochemical cells are produced and sold worldwide including 40 % of lithium-based batteries. The global transportation lithium-ion battery market in light duty vehicles will grow from $1.6 billion in 2012 to almost $22 billion in 2020 with a nominal battery pack cost down to $397/kWh. According to Frost and Sullivan, in 2009, lithium battery market made up 40 % of the global revenue, overall 37 % of income for Li-ion batteries and 3 % for primary batteries. Presently, the greater part of general-use LiBs is produced in China, Japan, and the Republic of Korea. From statistics of calendar year 2013 provided by METI, the total battery production by volume was 3.46 billion units that produced a value 683.4 billion yens [134]. This production included 61 % primary batteries and 39 % secondary batteries (Fig. 2.2).

Battery recycling is a necessary issue to avoid hazardous waste and pollution especially when automobiles will be electrically powered. The process could be similar to that used for lead-acid batteries. Estimation made by Chemetall [135] show that recycling could provide 50 % of the lithium requirement for new batteries by 2040. Presently, it costs about $1700 to recycle 1 ton of batteries of any chemistry and size. From calculations by Hsiao and Richter [136], a 100-Ah battery processed through recycling would return 169 kg of Li carbonate, 38 kg of Co and 201 kg of Ni assuming the LiNi_0.8Co_0.15Al_0.05O₂ (NCA) cathode chemistry, for a total value of the recovered materials that exceeds $5000.

2.6 Battery Modeling

The constraints encountered by electrified automobiles, HEVs, PHEVs, and EVs, lead to very complex powertrain architectures, which require an optimized energy management obtained by simulation of the batteries. Battery modeling can help to predict, and possibly extend its lifetime. Many battery models can be found in the literature that are electrochemical models and stochastic models; for a review see Jongerden and Haverkort [137, 138]. The open-circuit voltage (thermodynamic property) is the main parameter taken into account in the electrochemical model, in which the charge transfer overvoltage, the diffusion overvoltage and the ohmic drop are considered (Eqs. 1.6 and 1.7). Due to various nonlinear effects the expression of the lifetime t = Q/i does not hold for real battery, thus, a simple approximation for t under constant load i can be made with the Peukert’ law (Eq. 1.26). The equivalent circuit used for simulation is obtained from experimental data of electrochemical impedance spectroscopy (EIS) of the battery at given state-of-charge [139].

Electrochemical properties and battery management of lithium metal and lithium-ion cells were analyzed using several different models. The electrochemical model developed by Doyle et al. [140–142] consists of six coupled, nonlinear differential equations whose solutions obtained from the Fortran program named “Dualfoil” provide, with a very high accuracy, the voltage and current as functions of time and parameters of the battery electrolyte. Hageman has proposed the electrical-circuit models to evaluate Ni-Cd batteries [143]. A typical equivalent circuit includes five components: (1) a capacitor representing the capacity of the battery, (2) a discharge-rate normalizer that determines the lost capacity at high discharge currents, (3) a circuit to discharge the capacity of the battery, (4) a voltage versus state-of-charge lookup table, and (5) a resistor representing the battery’s resistance. The Rakhmatov model is an extension of the Peukert’s law based on the diffusion process of the active materials in the battery. To determine the lifetime one has to compute the diffusion process described by Fick’s laws. This analytical model applied to Li-ion cells does even better, with a 2.7 % maximum error and an average error of less than 1 %.

In the kinetic battery model (KiBaM) , Manwell and McGowan [144] used a chemical kinetics process with the battery charge distributed over two wells. For the calculation of the voltage during discharge, the battery is modeled as a voltage source in series with an internal resistance. The KiBaM can be used to model Li-phosphate-type Li-ion batteries because of the flat discharge profile of such a cell. The stochastic model based on discrete-time Markov chains was introduced by Chiasserini and Rao [145]. Modeling of Li-ion batteries has been obtained with a high accuracy, 1 % error. Later, in 2005, Rao et al. [146] proposed a stochastic battery model based on the analytical KiBaM. This simulation was successfully applied to AAA-type Ni-MH batteries with a maximum error of 2.65 %.

Tremblay and Dekkiche [147] presented an easy-to-use battery model applied to dynamic simulation software using the battery state-of-charge (SOC). Such a model is very similar to the Shepherd model that describes the electrochemical behavior of the battery directly in terms of voltage and current, often used in conjunction with the Peukert equation [148, 149]. However, contrary to the Shepherd model, it does not produce an algebraic loop. The battery is modeled using a simple controlled voltage source in series with a constant resistance, as shown in Fig. 1.7. This model assumes the same features for charge and discharge cycles. The open-voltage source is calculated with nonlinear Kalman filter that is an equation based on the actual SOC of the cell:

$$ E={E}_0-K\frac{Q}{Q- it}+A{e}^{\left(-B{\displaystyle \int idt}\right)}, $$

(2.21)

where E is the now-load voltage (V), E ₀ the battery constant voltage (V), K the polarization voltage (V), Q the battery capacity (Ah), A the exponential zone amplitude (V), B the exponential zone time constant inverse (Ah)⁻¹, and ${\displaystyle \int idt}$ the actual battery charge (Ah).

The long term calendar life of lithium ion cells for satellite and standby applications was studied by Broussely et al. [150]. In experiments, the capacity evolution was tracked as a function of storage temperature. Cells containing either LiCoO₂ or LiNi _x M _y O₂ positive electrodes coupled with a graphite negative electrode were float-charged at 3.8 or 3.9 V. This study focused on losses at the negative electrode and the data were fit to a model involving a rate-determining step governed by electronic conductivity of the solid electrolyte interphase (SEI) layer, following Arrhenius law as a function of temperature.