**1. Introduction**

The technological advancements that have taken place during the last few decades have created the need to store more energy in ever smaller volumes. Lithium-ion batteries can store large amounts of energy in small weights and volumes, making them the technology-of-choice for multiple applications. In addition to their use in everyday portable electronics and all-electric vehicles, lithium-ion batteries could also provide a way to store large amounts of energy harnessed by using solar cells and wind turbines [1]. It is clear that the market for lithium-ion batteries will continue to increase in the years to come [2,3]. However, urgent materials engineering advances are needed to enable the continued progress of the technology and answer the high demands of future applications.

The most widely used lithium-ion battery materials include oxides and phosphates for cathodes [4,5]. These materials are intercalation electrodes and thus have relatively low usable capacities of 100–150 mAh/g. Graphitic carbon is the most frequently used anode material, with spinel lithium titanate and elemental silicon gaining more interest: the titanate for its enhanced safety and elemental silicon for its huge maximum capacity of 3600 mAh/g, made possible by its alloying reaction with lithium [4,5]. Alloying anodes have presented challenges in battery fabrication, because large volume changes associated with the alloying reactions can lead to electrode pulverization and degradation. However, with recent advances in nanofabrication, these problems might soon be mitigated by the use of proper micro- and nanoscale battery constructions [6–10].

Whereas high capacity alternatives, such as silicon, are already being intensely studied to replace the present anode materials, oxide and phosphate materials still dominate the research on cathodes, despite their limited capacities. Fluorides were studied extensively in the 1960s and 1970s for use in primary lithium batteries (i.e., batteries that would not be rechargeable) due to their high theoretical capacities and energy densities [11]. It was hoped that these materials would act as conversion cathodes, forming lithium fluoride during discharge (Equation (1)):

$$n\text{Li}^+ + n\text{e}^- + \text{MF}\_{\text{ll}} \rightarrow n\text{LiF} + \text{M} \tag{1}$$

Conversion cathodes generally suffer from the same volume change problems as alloying anodes. However, by using the same nanofabrication methods as for anodes, these materials could provide a new, interesting class of lithium-ion battery electrode materials. Additionally, fluorides have also been reported to show high lithium-ion conductivities, making them possible solid electrolytes for all-solid-state thin film batteries [12–14].

Many enhancements in lithium-ion battery properties can be achieved by depositing component thin films onto 3D substrate geometries, as both active materials and protective layers on the active materials. During the 21st century, the thin film manufacturing method known as atomic layer deposition (ALD) has become a vital enabler of progressive technology nodes in microelectronics and is now becoming similarly important in the fabrication of Li-ion batteries. The advantages of ALD, including high film uniformity and excellent conformality over high-aspect-ratio substrates, make it ideal for the deposition of materials for ever-smaller, more complicated batteries: strict conformality is especially important for integrated, all-solid-state batteries, in which the small electrode thin film thicknesses can still produce high energy densities per footprint area when deposited into deep trenches [8]. The deposition of solid electrolyte materials, instead of using the current liquid electrolytes, could solve many of the safety issues associated with lithium-ion batteries.

This review briefly introduces the basic concept of lithium-ion batteries, some of the materials currently used in these batteries and the use of ALD in depositing these materials. Liquid electrolytes will not be discussed but instead potential inorganic solid electrolytes are shortly reviewed. Fluoride materials are presented as a potential "new" class of battery materials with uses as both electrodes and solid electrolytes for lithium-ion batteries. To motivate further studies on fluoride deposition using ALD, the literature in this area is reviewed.

## **2. The Lithium-Ion Battery**

## *2.1. Basic Principle*

Lithium-ion batteries are used for energy storage in applications ranging from cell phones and laptops to electric vehicles. The basic concept of a lithium-ion battery is the same as for any other battery: chemical energy stored in the electrodes is converted into electrical energy via a chemical reaction. Some of the battery types in use today are depicted in Figure 1. As can be seen, lithium-ion batteries have surpassed many of the older battery technologies in energy density. This is related to the fact that lithium ions are relatively small and light-weight, which makes it possible to obtain high energy densities from lithium-containing materials. In addition, lithium forms compounds with large enough potentials to produce so-called "high quality" energy, or high power densities [15]. The small size of the lithium-ion is also an advantage in that lithium-ions are highly mobile in many materials, ensuring only low energy losses due to kinetic effects.

Figure 2 depicts the basic schematic of a lithium-ion battery. The battery consists of two electrodes (the positive cathode and the negative anode), an electrolyte and two current collectors. Because the active materials in the battery, the electrodes, are separated by an inert electrolyte, the energy liberated in the chemical reaction between the electrode materials can be converted into electrical work. During battery operation, lithium-ions flow through the electrolyte from the anode to the cathode. At the same time electrons flow in the outer circuit from the anode to the cathode, balancing the net charge flow. These electrons can be used to do work, such as operate a laptop.

**Figure 1.** Comparison of energy densities in different battery types in use today. Reprinted with permission from [16]. Copyright 2001 Springer Nature.

**Figure 2.** Basic schematic of a lithium-ion battery, showing the flow of lithium ions and electrons during discharge.

In addition to the high energy and power densities, interest in lithium-ion batteries derives from the fact that most of them can be recharged multiple times without prohibitive loss of battery capacity. During recharging a voltage is applied to the outer circuit, which reverses the flow of electrons and Li+-ions, moving them from the cathode to the anode. During the charge-discharge cycling, the battery capacity should stay as constant as possible. Usually the capacity slowly degrades with increased cycling as a result of, for example, reactions between the electrolyte and the electrodes and changes in the morphology of the electrodes, such as breakage or delamination [15].

Most lithium-ion batteries employ an organic liquid electrolyte material, composed of organic carbonates such as ethylene, dimethyl and diethyl carbonates, lithium hexafluorophosphate and potentially different additives [5,15]. Liquid electrolytes enable very high lithium-ion conductivities, which are beneficial for the battery operation. At the same time, the electrolyte should be an electron insulator so that no self-discharge takes place [15]. Liquid electrolytes can accommodate volume changes in the electrodes during cycling, which reduces stress in the battery. However, liquid electrolytes are also a major reason for the safety concerns of lithium-ion batteries: the electrolyte can

decompose at the electrodes, most often on the anode and form a solid-electrolyte interphase layer, or SEI-layer [15]. Commonly the SEI-layer protects the electrodes from further reactions with the electrolyte but in the event of an incomplete SEI-formation, reactions can proceed further with pressure building up inside the battery. Since the electrolyte materials are flammable, explosions can occur [17].

#### *2.2. Conventional Electrode Materials*

The electrodes define the capacity and voltage of any battery. Since both lithium-ions and electrons move inside the electrodes, the materials need to be both ion and electron conductors. The reactions taking place in the electrodes can be roughly divided into conversion reactions, alloying reactions and insertion or intercalation reactions (Figure 3) [4]. Conversion and alloying electrodes generally provide higher battery capacities but there can be problems with electrode degradation during cycling due to repeated volume changes associated with the conversion reaction. Thus, conversion and alloying electrodes have more often been used in primary batteries, meaning non-rechargeable batteries.

**Figure 3.** Schematic illustrating the different kinds of reactions possible with lithium ions and electrode materials. (**a**) Intercalation; (**b**) Alloying; (**c**) Conversion. With alloying and conversion reactions, significant volume changes often take place in the electrode.

In rechargeable, or secondary, batteries intercalation reactions are the most commonly used, with the LiCoO2–C6 (graphite) battery being a prime example of this reaction. Even if intercalation electrodes do not generally suffer from volume changes of similar magnitude to conversion electrodes, they still require care in the charging-discharging process: too deep charge-discharge cycling can lead to irreversible changes in the electrode structure, which in turn can lead to a diminished battery capacity.

## 2.2.1. Cathodes

Intercalation cathode materials are most often lithium containing transition metal oxides, in which the insertion and extraction of lithium is made possible by the redox capabilities of the transition metal [4,18]. Although LiCoO2 is the most commonly used cathode material, manganese and nickel oxides are gaining interest as possible future cathodes [18,19]. Table 1 contains the potentials and capacities obtainable from some of the most studied cathode materials. Despite their common use in batteries today, transition metal oxide cathodes have some serious drawbacks. Firstly, the potentials obtainable from these materials are limited, which is a problem especially for high power density applications. Secondly, the capacities available from these materials are quite low as a result of the limited amount of lithium available before irreversible changes in the structures occur. For example, only half of the lithium-ions in LiCoO2 can be reversibly utilized [19]. Lastly, liquid electrolytes are known to give rise to dissolution of the transition metal, thus decreasing the cathode capacity even further [18]. However, thin film methods such as atomic layer deposition can be used to deposit thin layers (e.g., Al2O3 or AlF3) onto cathode materials to protect them from side reactions such as transition metal dissolution [20–24]. Some of these problems can also be circumvented by using mixture cathodes, such as LiNi(1−*y*−*<sup>z</sup>*)Mn*y*Co*z*O2, which can also produce slightly higher capacities [5,18,19]. In addition to oxides, sulphates and phosphates, such as LiFePO4, have also been studied extensively [19]. Despite being cheaper and less toxic than most cathode materials and providing a reasonable capacity, LiFePO4 suffers from low electronic conductivity, limiting its applicability [19]. For cathodes, alloying and conversion reactions have not gained much attention.

**Table 1.** Potentials and capacities obtainable from some of the most studied lithium-ion battery cathode materials.


#### 2.2.2. Anodes

Table 2 presents some potential anode materials for lithium-ion batteries. For an anode, a potential as low as possible is desired to reach a high cell voltage. As with the cathode materials, intercalation anodes are more common than the other types of anodes. However, a lot of work is now being put into studying materials such as silicon and tin as lithium-ion battery alloying anodes [4,5,9,25].

**Table 2.** Potentials and capacities obtainable from some of the most studied lithium-ion battery anode materials.


Metallic lithium was, unsurprisingly, the first choice for the anode of a lithium-ion battery [18,26]. However, safety concerns such as dendrite formation [26] have moved the interest towards other anodes such as graphitic carbon C6, which produces capacities of 370 mAh/g [5,15]. The use of C6 still does not help avoid dendritic lithium deposition due to the very low electrode potential of carbon anodes. One possible replacement for carbon is the spinel lithium titanate Li4Ti5O12, which has a reasonable capacity with a very small volume expansion during lithiation. In addition, the titanate is environmentally benign with a reasonable cost [28]. A major drawback of this material and many insertion oxide anodes in general, is the high electrode potential, which results in cells with low operation voltages [5,28].

Of the alloying anodes, elemental silicon has attracted much attention because of its low cost, abundance in nature and high specific capacity of over 3500 mAh/g [28]. However, alloying the maximum 3.75 lithium-ions per one Si atom produces volume changes of up to 300% in the electrode, limiting the cycling ability of the anode [5]. Nanostructured silicon is now studied in the hope of

alleviating the problems with volume expansion [5,9]. In addition, atomic layer deposition has been utilized in contact with Si anodes to improve their mechanical integrity and to stabilize the interfaces of the anode [10].

Similarly, elemental tin has attracted much attention but suffers from the same problems as silicon [9]. Attempts have been made to circumvent the volume change problems by using conversion anodes composed of, for example SnO or SnO2. In these anodes, lithium first forms lithium oxide and tin is reduced to metallic tin, which can further alloy lithium and is responsible for the reversible capacity [28]. However, the problems related to volume changes still persist to some extent in these anodes [28]. Many transition metal oxides have also been studied as conversion anodes, with capacities ranging from 600 to 700 mAh/g [27]. In addition to the persisting volume change problem, conversion anodes often also suffer from large overpotentials [9].

## *2.3. Conventional Solid Electrolyte Materials*

Solid electrolyte materials are the enablers of all-solid-state Li-ion batteries. These materials have stringent property demands: they must be unreactive at the electrode potentials and have a high lithium-ion conductivity at room temperature [15,30]. In addition, they must be good electron insulators to avoid self-discharge and short-circuits [4]. Despite the high specification for material properties, a vast number of solid electrolyte materials suitable for lithium-ion batteries have been reported in the literature [30–33]. In addition to inorganic ceramics, composites and polymer mixtures can be used as solid electrolytes [31,33,34].

Inorganic fast lithium-ion conducting materials can be single-crystalline, polycrystalline or amorphous, with many different structural types reported for the crystalline materials (Table 3, Figure 4) [31,32]. Generally, an amorphous electrolyte material would be preferred as grain boundaries in crystalline materials can lead to both impeded ion movement and electron leakage and thus poorer insulating properties [33,34]. In addition, amorphous materials can provide isotropic lithium-ion conduction in a wide range of different compositions [32].

**Table 3.** Structures and ionic conductivities of some of the most studied solid inorganic lithium-ion conducting materials.


Amorphous, or glassy, electrolytes can be roughly divided into oxide and sulphide glasses [32]. Probably the most studied amorphous solid electrolyte material is nitrogen doped lithium phosphate (lithium phosphorus oxynitride), LiPON, which is in fact already in use in thin film lithium-ion batteries [32,33]. The material is usually deposited by sputtering in a nitrogen atmosphere and the resulting films have lithium-ion conductivities of the order of 10−8–10−<sup>6</sup> S/cm [32], as compared to the ~10−<sup>2</sup> S/cm for liquid electrolytes [5]. The success of LiPON comes not only from its reasonably high lithium-ion conductivity but also from its excellent stability against the electrode materials [32]. Sulphide glasses, on the other hand, have been studied much less than the corresponding oxides, mostly because of their reactivity in air and corrosiveness [32].

**Figure 4.** Bulk ionic conductivities of selected oxides and phosphates as a function of temperature. Reprinted with permission from [33]. Copyright 2009 Elsevier.

Out of crystalline electrolyte materials, the perovskite (Li,La)TiO3 (LLT) is known to show high bulk conductivity. However, reduction of Ti4+ and the consequent increase in electronic conduction are problems associated with this material [31,33]. Titanium-free perovskites, such as (Li,La)NbO3 have thus been studied [32]. Phosphates with the NASICON (sodium superionic conductor) structure are also known to show high lithium-ion conductivities [32]. In Table 3 LiTi2(PO4)3 is given as an example but many other +IV oxidation state metals can be substituted for titanium in the structure. LISICON (lithium superionic conductor) materials, on the other hand, are mixtures of lithium silicates or germanates, lithium phosphates or vanadates and lithium sulphates and can produce similar conductivities as the NASICON structures [32]. In addition to these material classes, oxides such as Li6BaLa2Ta2O12 with the garnet structure have been studied extensively and shown to have reasonable ionic conductivities [31]. Another benefit of the garnet materials is their high chemical stability in contact with the electrodes [32].
