Lithium-ion battery operation

Lithium-ion battery operation

CHAPTER Lithium-ion battery operation 3 Chapter outline 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 Where is the lithium in lithium-ion? ...

5MB Sizes 0 Downloads 8 Views

CHAPTER

Lithium-ion battery operation

3

Chapter outline 3.1 3.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9

Where is the lithium in lithium-ion? ................................................................... 44 How do the anode and cathode work together? ................................................... 45 Reduction/oxidation (redox) process ................................................................... 48 Intercalation ..................................................................................................... 51 Cations and anions ............................................................................................ 52 Solid electrolyte interphase ............................................................................... 53 What does “nano” mean? .................................................................................. 56 Thermodynamics ............................................................................................... 59 Failures modes ................................................................................................. 60 3.9.1 Internal short circuit ........................................................................63 3.9.2 External short circuit ........................................................................66 3.9.3 Thermal runaway .............................................................................67 3.9.4 Cascading failure .............................................................................68 3.9.5 Impact and effects of temperature on cell aging .................................70 3.9.6 Impedance ......................................................................................71 3.9.7 What happens during overcharge .......................................................71 3.9.8 What happens during overdischarge ..................................................72 3.9.9 Influence of impact, crush, and penetration .......................................74 3.9.10 Aging mechanisms .........................................................................74

I was recently asked a good question “why lithium”? Lithium is the lightest metal on the periodic table with an atomic number of three, with only the gases helium and hydrogen having lower atomic weights (two and one, respectively). Being one of the alkali earth metals means that lithium is also highly reactive. This combination of low weight and high reactivity means that lithium contains more energy per unit of weight than other metals which is what makes it such a good element to use for energy storage. In comparison, nickel has an atomic number of 28 and lead has an atomic number of 82. Remember that the atomic number represents the number of protons in the nucleus of the element. These higher atomic numbers correspond to the differences in energy density in electrochemical cells of each. Lead acid batteries, with lead’s atomic number of 82, offer on average about 40 Wh/kg cell energy density and nickel metal hydride, with nickel’s atomic number of 28, falls in between lead and lithium with average cell energy density of about 80 Wh/kg. Lithium-Ion Battery Chemistries. https://doi.org/10.1016/B978-0-12-814778-8.00003-X # 2019 Elsevier Inc. All rights reserved.

43

44

CHAPTER 3 Lithium-ion battery operation

FIG. 23 Lithium atom.

Another reason lithium is particularly well suited for electrochemical energy storage is that it has one valence electron. You may remember from Chapter 2 valence electrons are the ones that are “mobile” and can be lost or gained the easiest. The lithium atom contains three electrons that circle around a nucleus containing three protons and three neutrons. These three electrons orbit around the nucleus in two energy bands. The electrons in the innermost band are not mobile, but the electron in the outer band is mobile making it a valence electron. There is also a “hole” in that outer energy band which can accept an electron. By accepting an electron it would have one more electron than proton giving it a negative charge and making it an anion. However, if that single outer electron is removed it has one more proton than electron making it positively charged and a cation (Fig. 23).

3.1 Where is the lithium in lithium-ion? This is another question that comes up quite often, where is the lithium in a lithiumion battery? The question came up a lot right after Canada issued a ban on the air transport of lithium metal batteries in 2016 (Roy, 2016). In this case they were concerned mainly about primary, or nonrechargeable, lithium-ion batteries that used a lithium-metal anode because in its metallic form lithium metal is known to be quite reactive to oxygen and water. Lithium metal has been used in anodes for many years in primary batteries but until recently has been limited in use in secondary batteries because lithium metal is extremely reactive to both oxygen and water. However, in secondary, or rechargeable, lithium-ion batteries there is no lithium metal at least as of today. Future solid-state batteries may use lithium metal as anode

3.2 How do the anode and cathode work together?

electrodes again but today, instead of being used as the metal anode the lithium-ions are mixed into the electrolyte in the form of lithium salts with most of it going into the cathode active materials (Dahn & Ehrlich, 2011). During the initial formation process the lithium is forced out of the cathode and through the electrolyte and separator and into the negative anode during a controlled charge and then back into the cathode during a controlled discharge. It is this process that creates the solid electrolyte interface (SEI) layer, which is essentially an electrochemical process that creates a coating on the anode materials that is made up of lithium-ions that form permanent bonds with the molecules in the anode. This results in some loss of lithium during these first cycles and a permanent loss of initial capacity as the SEI layer is formed. The lithium used in lithium-ion batteries is positively charged cations. During cycling the positively charged lithium-ions are forced by way of the redox process out of the active material on the anode through the electrolyte and separator and are reduced at the cathode. The process is then repeated with the lithium-ions traveling from anode to cathode and back again. Remember that because negative and positive particles are attracted to each other, the lithium cation will always be pulled toward the negative electrode which, as we talked about in the previous chapter, changes during operation. We will review the redox process in more detail in an upcoming section of this chapter. Some cell manufacturers are beginning to use a process called prelithiation wherein lithium in its powder form is added into the anode during electrode processing. This acts to pump up the amount of lithium in the cell and increasing the energy density by storing some active lithium in the negative electrode prior to charge/discharge cycling. This helps to reduce the first cycle loss due to the lithium being consumed in the SEI formation process and lithium that is lost due to the aging processes. And while most prelithiation is mainly being done to the anode, but it is also being done to the cathode side by “over-lithiating” the cathode (Holtstiege, B€armann, N€ olle, Winter, & Placke, 2018). However, there are still some question as to the impact on cycle life when you overlithiate the anodes in this manner.

3.2 How do the anode and cathode work together? The anode and cathode work together like opposite poles of a magnet, sending the lithium-ions back and forth between them each time releasing electrons which create the current flow. In fact, when a metal electrode is immersed in an electrolyte, the electronic charge on the metal attracts ions of the opposite charge and orients the magnetic poles (dipoles) of the elements in the solvent. There is a layer of charge in the metal electrode and a layer of charge in the electrolyte. This charge separation establishes what is commonly known as the “electrical double layer” between the electrodes (Salomon, 2011, p. 2.10). But this small amount of magnetic attraction is only one of the driving forces moving the lithium-ions back and forth and generating current.

45

46

CHAPTER 3 Lithium-ion battery operation

FIG. 24 Active material with lithium-ions inserted.

The basic function of the anode and cathode, the negative and positive electrodes, is to act kind of like a lattice framework to hold the lithium-ions as they pass back and forth. The lithium-ions fill into the gaps within the active materials similar to what is shown in the example in Fig. 24. In this over-simplified example imagine that the large, dark colored spheres represent the “active material” either anode or cathode material. The silver spheres represent the lithium-ions so in this example we see them filling in the voids between the active materials. When the silver colored lithium-ions are moved from anode to cathode during discharging those ions travel from one side of the electrode to the other and attach to the cathode material surface. This process of attracting the ions to the surface of the active material is known as adsorption or intercalation. Adsorption, different from absorption, is the process when atoms or ions get attracted to the outside surface of a material and collect on the surface of the material rather than entering or diffusing into the material (Julita, 2010). In the case of adsorption reaction of a molecule in a lithium-ion cell, the adsorbed species, lithium-ions in this instance, occupy vacant sites on the electrode surface, thus partly blocking it as shown in the example in Fig. 24 (Lefrou, Fabry, & Poignet, 2012, p. 66). While these electrodes may seem to be very simple solid pieces, they are not quite as simple as they would appear at first glance. Each electrode is a composite body made of a combination of active material, binder, performance enhancing additives, and conductive filler and a conductive substrate (Salomon, 2011, p. 2.1).

3.2 How do the anode and cathode work together?

The conductive substrates, most often copper and aluminum, are also coated with a conductive material, usually a form of carbon, that helps bind the active materials to the substrate. As a rule, cell designers prefer to design electrodes with large surface areas since this will act to minimize the energy loss due to the activation and concentration polarizations at the electrode surface that occur as the SEI layer is formed and to increase the electrode efficiency or utilization. This is most often accomplished by using a porous electrode design. A porous electrode consists of porous matrices of solid and void spaces. The electrolyte penetrates the void spaces of the porous matrix. A porous electrode can provide an interfacial area per unit volume several times higher than that of a planar (flat) electrode. What this all means is that a high surface area allows for more electrolyte to penetrate it and it is through the electrolyte that the lithium-ions can travel into the active material thus storing more energy. Reactions at an electrode surface are characterized by both chemical and electrical changes. Electrode reactions can be as simple as the reduction of a metal ion and incorporation of the resultant atom onto or into the electrode structure (Salomon, 2011, p. 2.4). But in this type of active porous mass, the mass transfer condition in combination with the electrochemical reactions that occur at the interface can also be very complicated (Salomon, 2011, p. 2.18). The reactive interface surfaces are those where at least one chemical reaction occurs (as shown in Fig. 25) and the current circulation through an interface involves one or more electrochemical reactions, which tend to be heterogeneous in nature. These chemical reactions include, of course, the reduction and oxidation processes. However, depending on things like rate and temperature other chemical reactions could take place and these are typically called “side reactions” as they are not the intended reactions. Studying the reactions

FIG. 25 Interfacial areas.

47

48

CHAPTER 3 Lithium-ion battery operation

that occur at these interfaces is extremely important in order to ensure high efficiency and long life of a lithium-ion battery. The more efficient the reactions at the interfaces the better the battery performance will be. The movement of the lithium-ions is referred to as mass transport and can be classified into three categories. The first is called migration which is the movement of charged ions due to an electric field. The second mass transport process is diffusion which is the movement of ions that are submitted to a concentration gradient, in this case the ions diffuse from areas of greater concentration to areas of less concentration. A diffusion layer is generally next to the interface layer, which may also see the other two forms of mass transport, migration and convection. The thickness of the diffusion layer is generally only a few nanometers (μm) thick. The third type of mass transport is from convection, which is the movement of the medium when it is in a liquid form (Lefrou et al., 2012, pp. 61–62). These processes that occur during mass transport where cations and anions are transported to and from the electrode surface are perhaps the most important of all processes that occur in a lithium-ion cell because without them, none of the other process can take place (Breitkopf & Swider-Lyons, 2017).

3.3 Reduction/oxidation (redox) process The next process that we must understand in electrochemical lithium-ion cells, after adsorption, is what is referred to as the redox process. This is the reduction and oxidation of ions. Historically the term oxidation referred to an element that reacted with oxygen and reduction referred to a metal that was produced from its mineral ore (Matson & Orbeak, 2013, p. 68) but today these refer to the reduction and oxidation processes that occur as the lithium-ions are transferred from anode to cathode and back. It sounds a bit backwards, but if the atom gains electrons it is said to have been reduced and if the atom loses an electron it is said to have been oxidized. It is important to note that the reduction reaction always takes place at the cathode and the oxidation reaction always takes place at the anode. But to those of us that are not chemists this is a very confusing statement because we think of the cathode as being the positive active material such as NMC or LFP and the anode as being the negative active material, usually graphite. But in actual operation we end up with a cathodic reaction (reduction) that takes place on one electrode and an anodic reaction (oxidation) that takes place on the other. During charging the cathodic reaction takes place in what we refer to as the anode since it oxidizes and releases electrons while the anodic reaction takes place in the cathode since it reduces and gains electrons. When the cells switch to discharging the cathodic reaction takes place at what we would consider the cathode as it is here that the electrons are gained as the electrons follow the circuit from the anode side where an anodic oxidation reaction occurs to release these electrons (Fig. 26). If these reactions did not change between electrodes like this then the cells may not be able to recharge; once the electrons were spent the cell would be dead and quite useless. This gets quite confusing once we start talking

ReDox operations.

3.3 Reduction/oxidation (redox) process

FIG. 26

49

50

CHAPTER 3 Lithium-ion battery operation

about the active materials, so moving forward we will use the term cathode to refer to the electrode with the active cathode materials such as NMC, NCA, LFP, LMO, or LCO and we will use the term anode to refer to the electrode with the active materials of graphite, carbon, silicon, or titanium. Reduction always takes place at the cathode and oxidation always takes place at the anode… but anode and cathode change during charging and discharging When current flows through an electrochemical cell one of the electrodes becomes the anode and the other the cathode as stated above, but the link between the voltage of an electrode and its role as an anode or cathode is not always constant. As we talked about previously, depending on the operating conditions of the system, charging or discharging, an electrode can be either the anode or the cathode and will change its voltage potential accordingly. The current is positive if the electrode is on the oxidation site, the anode. The current is negative if the electrode is on the overall reduction site, the cathode (Lefrou et al., 2012, pp. 27–28). Oxidation occurs when elements combine with other elements that are more electronegative and reduction occurs when an element combines with another element that is less electronegative. Remember that the lithium-ion is a cation, positively charged, so it will always be pulled toward the more electronegative electrode. Oxidation can occur in one of three ways, either gaining oxygen, losing hydrogen, or losing an electron and reduction occurs in the opposite manner, losing oxygen, gaining hydrogen, or gaining an electron. The reduction/oxidation processes are shown together since if an atom loses an electron the other must gain it to ensure the balance is maintained. This is what causes the cathode and anode to “swap” during operation; once all of the free valence electrons have been released from the anode during discharge they flow through the current collector through the electrical machine and back into the other electrode, the cathode where the reduction reaction occurs. Therefore when the free electrons are all released from the anode the cell is discharged. During charging the opposite reactions occur; forcing a current into the cathode releases the electrons which pass through the circuit and into the anode material. These movements of the lithium-ions are also sometimes referred to as the “rocking chair” effect or as “shuttling” of lithium-ions due to the back and forth nature of the reactions. The simple figures later show how the rocking chair picks up the lithium-ions at the anode and then as the chair rocks forward drops off the lithium-ions at the cathode (Fig. 27). The other term that you may hear frequently is that of mass transport, which is simply the movement of any element, atom, or ion that has mass through some form of medium. Mass transport to or from an electrode can occur in one of three different ways: (1) through convection and stirring, (2) through electrical migration in a voltage potential gradient, and (3) through diffusion in a concentration gradient (Salomon, 2011, p. 2.15). Diffusion processes are typically the mass-transfer

3.4 Intercalation

FIG. 27 Lithium-ion rocking chair effect.

processes at work in the majority of battery system where the transport of species to and from reaction sites is required for maintenance of current flow (Salomon, 2011, p. 2.16). It is the redox reaction and that movement of electrons that generates the electric field and forces the current flow in the battery cell. The electric current produced between two electrodes can be determined by looking at how many volts are produced from the transfer of electrons. When you measure this voltage, you can determine how many volts are required to reduce or oxidize an element (Matson & Orbeak, 2013, p. 75). In the lithium-ion battery you want the cathode to have the more positive reduction potential since cell reduction occurs in the cathode and oxidation occurs in the anode. The number of electrons that can be moved is determined by how many “empty holes” are left in the valence shell of the atom/ion. Referring back to Chapter 2 we saw that the lithium atom carries three electrons, one of which is a valence electron. It is the valence electron that determines how many electrons the ion can transport. In the case of an electrically neutral lithium atom, we see that in the valence shell, or outer shell, there is only the one electron; however, there are seven empty holes in that second energy shell that can be filled with electrons during charging and discharging as shown in Fig. 28. The next factor that affects how many electrons an atom can carry is the diameter of the nucleus, the smaller the atomic radius the fewer electrons it has room to hold. Lithium has an atomic radius of 151.9 pico-meters (pm).

3.4 Intercalation At the beginning and end of the RedOx process the lithium-ions get inserted into the anode or cathode in a process called intercalation. This rather confusing looking term that simply means that the lithium-ions are inserted or removed from a host material. Since the lithium-ion is such a small particle it easily intercalates into the much larger

51

52

CHAPTER 3 Lithium-ion battery operation

FIG. 28 Valence electrons and holes in valence shell.

compound materials. This is a reversible process, which allows the repeating charge and discharge of the cell. The lithium-ion must also be inserted without creating significant structural changes in the host materials otherwise it will damage the host material and reduce the cycle life. In the case of lithium-ion batteries the cathode active materials and the anode active materials are the hosts. The metal oxide materials in the cathode have either a layered- or tunnel-type crystal formation which allows the lithium-ions to be inserted in between the layers or into the tunnels. The crystal structure of the carbon-based anodes also has a layered structure which allows the lithium-ions to be inserted in between the layers of active material (Dahn & Ehrlich, 2011).

3.5 Cations and anions Building on the last section and as we discussed in Chapter 2 when an atom carries excess negatively charged electrons the atom becomes a negative ion, otherwise called an anion. When an atom loses its negatively charged electrons it becomes a cation. It is really a lithium cation that is moving from the anode to the cathode and back in the lithium-ion cell. During these reactions, the ions create a temporary bond with the active materials, this is a form of ionic bonding which occurs when atoms donate or receive electrons rather than share them (Matson & Orbeak, 2013, p. 68). Therefore, a lithium-ion is simply a lithium atom that carries either a positive or negative electric charge. Using the lithium atom as an example, we know that it has three protons in the nucleus which are positively charged. We also know that it has three electrons, which are negatively charged. Therefore in its base condition it is

3.6 Solid electrolyte interphase

FIG. 29 Lithium cation and anion.

electrically neutral (3 “P +” and 3 “E ”). It becomes an ion, or a cation, by giving up one of its electrons. Now it has three protons and only two electrons giving it a net positive charge of plus one (3 “P+” and 2 “E ”). On the other side of the equation it becomes an ion, or an anion, by gaining one (or more) electrons. Now it has three protons and four electrons giving it a net negative charge of minus one (3 “P +” and 4 “E ”) as shown in Fig. 29. In the case of the migration form of mass transport discussed earlier, negatively charged anions and positively charged cations move in opposite directions. Positive charges migrate in the same direction as the electric current and negative charges move in the opposite direction as shown in Fig. 30. The directions that the ions move are linked to the direction of the current flow. Simply stated the positively charged cation is always drawn to the more electronegative electrode. In general then we can say that anions migrate toward the cathode and cations migrate toward the anode (Matson & Orbeak, 2013, pp. 43–46). But remember that while I am using a lithium cation and ion in my image here, the lithium-ion is always a cation. But there may be both cations and anions of other species and elements that are in the active materials, electrolyte, and salts.

3.6 Solid electrolyte interphase During the cell formation process most lithium-ion chemistries undergo a chemical reaction that generates a protective layer at the interface of the anode material and the electrolyte which is known as the SEI layer. The growth of the SEI layer is a result of

53

54

CHAPTER 3 Lithium-ion battery operation

FIG. 30 Cation and anion movement during charging.

the chemical reaction between the anode and the electrolyte solvents that is created as lithium-ions from the cathode and the salts effectively get “stuck” in the surface layer of the anode. This is a bit of a simplification as there is much more than just lithium getting stuck at the surface, in fact the SEI layer consists of a combination of materials including reduced or decomposed solvents in the electrolyte, salts used in the electrolytes, lithium-ions, and the impurities in the electrolyte (An et al., 2016). Once formed, the SEI layer “impedes” the flow of lithium-ion between the anode and cathode. The length of time that the cells are in the formation process will determine how thick the SEI layer forms. The thicker the SEI layer the greater the impedance to the lithium-ion flow which ultimately will reduce cycle life. The SEI layer is formed during the first charge/discharge cycle that the cell undergoes during what is known as the formation process. The specific charge rate and amount of charge is different for most cell manufacturers due to the differences in their chemistries and in fact is often a pretty highly guarded secret. However, in most cases the first cycle is generally a very slow charge and equally slow discharge, usually several hours (or at very low C-rates). The speed at which the SEI layer is formed determines the thickness of the layer and will impact the capacity of the cell. In general, a first cycle charge rate of 0.05C up to 0.2C is preferred to create the best SEI layer. Higher C-rates in the first cycle tend to reduce capacity and form a less stable SEI layer. By less stable, I mean that the SEI layer may be thinner and may be susceptible to crack and will then have to recreate new SEI to fill the cracks which

3.6 Solid electrolyte interphase

will consume free lithium and thus reduce the cell capacity. Therefore SEI development is very closely monitored at all cell manufacturers and is part of their “secret sauce” for manufacturing the cells. But it is important to note that in most cases the SEI will continue to develop during subsequent formation cycles at different rates and different temperatures until it is generally fixed. This is the reason many cell manufacturers go through two formation cycles. The first is to create the SEI layer, which is followed by a degassing operation, the second formation cycle occurs after the degassing operation and helps to set the final SEI thickness and determine the capacity, voltage, and impedance of the cell (An et al., 2016). Another important point is that during this first cycle, the cell may lose up to 10% of its original capacity as the lithium is consumed in the creation of the SEI layer. This is referred to as irreversible capacity loss as it can never be regained (An et al., 2016). Most of us tend to use a very simplistic view of the cell materials and view the active materials as a solid mass with the SEI layer forming right on the surface of that mass. The active materials are not really a solid mass as shown in Fig. 31 which shows the different components of the cell, but instead the active material is a composite material made of compressed molecules. The SEI layer is a small layer that forms on the surface of the molecules as shown in Fig. 32. In the following image you see a thin layer that forms on the surface of those molecules that are exposed to the electrolyte interface layer. It does not form throughout the entire material thickness. One of the most unique properties of the SEI layer is that it is both electrically insulating but still permeable enough to be ionically conductive; it should also be no more than a few angstroms thick (a few tenths of a nanometer), have high strength, be tolerant to expansion and contracting during operation; it should be insoluble to the electrolyte after formation; and it should be stable over a wide range of temperatures

FIG. 31 Cross section of lithium-ion cell.

55

56

CHAPTER 3 Lithium-ion battery operation

FIG. 32 SEI layer.

and voltage potentials (An et al., 2016; Huggins, 2009). This means that it provides a level of protection against internal short circuit to some minor extent, but still allows the lithium-ions to pass through it. The SEI must also entirely cover the surface area of the anode material to prevent continuous decomposition of the electrolyte that was not initially protected (An et al., 2016). If not completely covered the cell will continue to lose lithium during the cycling process and gases may continue to form as the lithium is consumed to continue to create SEI—both of which will drive the cell into an early failure. It is also important to note that the during the SEI layer formation process there are frequently some amount of gases that are generated as a by-product of the chemical reactions occurring. This is why lithium-ion cells need to go through a degassing process after the SEI formation period and prior to final sealing of the cells. This process allows these gases to be removed through a vacuum process to prevent them from causing side reactions of reducing the life of the cell.

3.7 What does “nano” mean? The term “nano” is used a lot today, especially when referring to advanced technologies such as batteries. But what does it really mean? The name comes from the ancient Greek word nanos, meaning “dwarf.” But in simple terms it just means something that is really, really small. Nano is short for nanometer, which is a unit of measurement representing one billionth of a meter (0.000000001 m). The typical range of sizes for something to be considered a “nano” is between 1 and 100 nm. When it is used in terms like nanoscience and nanotechnology it is referring to phenomena that occur on the atomic scale. To put it in perspective the deoxyribonucleic acid (DNA) found in the human body has a diameter of about 2.2–2.6 nm while a lithium atom has

3.7 What does “nano” mean?

diameter of about 0.3 nm and the electrons within those atoms have a diameter of only 0.000001 nm. So your DNA is two hundred and sixty million percent (260,000,000%) larger than the electrons we are dealing with in the lithium-ion battery, it is even 867% bigger than the lithium atoms. In his introduction to his book “Nanomaterials for Lithium-ion Batteries” (Yazami, 2014) Rachid Yazami describes the benefits of nanomaterials as offering both increases in energy and power density as compared to micron-scale technologies. These ultra-small materials can provide additional sites for storing lithium due to the high ratio of surface area to volume. Additionally, nanomaterials are believed to be able to fill those sites much faster than other technologies. Both features offer the benefit of increasing energy and power density. Nanomaterials can be applied to the cathode, anode, separator, or other components. Specific to advanced lithium-ion batteries there are many different types of nanomaterials, including nanotubes, nanorods, nanowires, nanoflakes, nanoparticles, nanosheets, nanoshells, nanoribbons, and nanospheres. Researchers are developing many unique form factors of nanomaterials to achieve different characteristics. Nanomaterials fall into two major categories based on the types of materials that are used such as carbon nanotubes (CNT) or silicon nanotubes (SNT). A simplified view of a nanowire may look like grass growing on your lawn as shown in Fig. 33. The nanostructures are typically grown directly on the metal current collector. A couple of important things to note in this image, first is that each of the nanowires is directly connected to the current collector at the bottom of the image which offers several benefits such as allowing for the volume changes in the horizontal direction that occurs during the lithium insertion process as shown in Fig. 34, which is especially helpful with silicon and tin anodes. This direct contact structure also has the potential to eliminate the need for binders and conductive additives that are used in standard anode materials today. You can see in this second image that the nanowires get “fatter” after the lithium-ions are inserted. This is what enables the greater energy storage of the nanowires, the nanowires allow all of the nanoactive materials to join in the charge storage process which increases the energy density, and due to

FIG. 33 Nanotubes on a current collector.

57

58

CHAPTER 3 Lithium-ion battery operation

FIG. 34 Nanotubes after lithium insertion.

the single-dimension electric pathways they allow for good charge transport (Sun & Chang, 2017, p. 866; Yazami, 2014, p. 5). Also keep in mind that due to the nanowire structure all the nanowire surfaces are directly in contact with the electrolyte rather than just the surface which again assists in the charge transport and energy density benefits. The nanotube only differs from the nanowire in that it is grown with a hollow structure such that it looks like a tube. The benefit of this is that it increases the surface area that contacts the electrolyte. Both the inside and the outside surfaces of the tubes contact the electrolyte which improves the ion transport capability. The carbon nanotube can be categorized into several different categories including single-wall carbon nanotubes (SWCNT) and multiwall carbon nanotubes (MWCNT). This differentiation is based on the thickness of the CNT and the number of coaxial layers of the nanotube structure (Goriparti et al., 2014). Due to the extremely small sizes of these materials the only way to get a real image of them is by using a scanning electron microscope (SEM) or a transmission electron microscope (TEM). The following images reprinted with permission from Springer Nature in the Nature Nanotechnology journal article “Stable cycling of double-walled silicon nanotube battery anodes through solid–electrolyte interphase control” by Wu, Chan, Choi, Ryu, Yao, McDowell, Lee, Jackson, Yang, Hu & Cui (2012) show good examples of double-walled carbon nanotube structure using a SEM microscope and a TEM microscope (Fig. 35). Nanowires are typically manufactured through several different processes including by a chemical vapor deposition (CVD) process or an electrodeposition (ED) process. From a manufacturing perspective, the biggest benefit of using either CVD or ED is that it simplifies the electrode manufacturing process significantly. Using vapor deposition or electrodeposition eliminates the need to mix the electrode materials with binders and solvents to create a slurry which is then pasted onto the current collector foils. With ED and CVD, the active materials are effectively grown directly on the current collector. These processes may significantly reduce the manufacturing cost as these technologies are scaled up into large-scale production capability (Sun & Chang, 2017). Electrodeposition and similar processes have been used in

3.8 Thermodynamics

FIG. 35 Scanning electron microscope image of nanowire structures showing a DWCNT using an SEM microscope in images (A) and (B) and with a TEM microscope in image (C).

high-volume manufacturing for many years, including in batteries for NiMh electrode coatings. So there is some precedence set for the use of these types of processes in lithium-ion battery manufacturing. But these technologies may still not yet be ready for the high-volume manufacturing required for mass manufacturing in lithium-ion batteries. The cost of making nanomaterials is still relatively high, so there is work to be done to improve the processes and reduce the costs. They also struggle from the perspective of making a thick material as these processes are typically limited to the nanometer-scale thickness. Today what many developers are doing is creating the nanomaterials as either a layer that the active materials are coated onto or as an additive material that can be added in with the active materials to improve their conductivity and performance. One final comment on nanomaterials is that there are some special precautions that need to be taken from a human perspective. Since these materials are so small the traditional personal protective equipment (PPE) will not provide any protection from the ingestion of these materials or from absorption through the skin. Therefore it is important to use specialty PPE when handling and working with nanomaterials.

3.8 Thermodynamics The science of thermodynamics is concerned with accomplishing “work.” But this does not mean that we are engaged in thermodynamics at the office from nine to five every day, rather thermodynamics describes either mechanical work, like the use of a wheel or lever, or in the case of batteries, electrical work. In the electrochemical lithium-ion cell, thermodynamics is specifically concerned with the maximum

59

60

CHAPTER 3 Lithium-ion battery operation

amount of work (power) that a reaction can deliver and how much work (power) is required to make that reaction occur (Breitkopf & Swider-Lyons, 2017). As mentioned earlier, it is the flow of electrons during the redox process that creates the flow of electrical current and it is the external power applied to the circuit that provides the electromotive force to move the ions back and forth between the electrodes. It is important to note that thermodynamic laws apply equally to mechanical power and electrical power. The first law of thermodynamics states that energy can neither be created nor destroyed and only the form in which the energy exists can be changed. The second law of thermodynamics states that heat cannot be completely converted into another form of energy and that all energy transformation processes are irreversible and occur in a preferred direction (Robert Bosch GmbH, 2007). Thermodynamics is an important aspect of the lithium-ion battery operation as it describes the properties of the chemicals and solutions and the impact of temperature and pressure during the reactions. It also helps to define and describe the interaction of the lithium-ion chemistry with the environment. Thermodynamics in batteries also differs somewhat when compared to other chemical reactions due to the electric field and the electrodes. In common chemical reactions outside of the electrochemical battery, energy is delivered as heat while no work is exchanged. Thermodynamics is also useful in defining and determining the losses that occur during the chemical reactions within the cells (Hebecker, 2017). In the lithium-ion battery, the polarization effects that occur as part of the redox process consume part of the energy of the cell as it is converted into waste heat, so not all of the theoretically available energy stored in the electrodes can be fully converted into useful electrical energy (Salomon, 2011, p. 2.1). When energy is no longer able to perform work, such as when electrical energy is converted into heat, it is described as entropy (Robert Bosch GmbH, 2007). One other term that falls within the realm of thermodynamics which I want to briefly discuss is the term adiabatic. This is frequently discussed in terms of looking at the thermal efficiency of a complete battery system. But it applies in cell development as well, especially during testing, cell development, and characterization. Simply stated, a system is described as being adiabatic when heat is neither supplied nor is dissipated (Robert Bosch GmbH, 2007). In measuring a cell or battery’s heat generation capabilities, an adiabatic test might be used wherein the cell is insulated and the test chamber temperature is matched with the cell, or pack, temperature such that the cell does not gain any heat from the environment but also does not release any heat to the environment. This can be a useful way to map the thermal characteristics of a cell.

3.9 Failures modes One of the most important aspects of battery design, whether at the cell or pack level, is safety. Safety can come in the form of changes we make to the chemistry within the cell, to the separator used, to the active material selection, and to the electrolytes

3.9 Failures modes

FIG. 36 Safe operating zones for temperature and voltage.

used. But ultimately and in all instances lithium-ion safety is concerned with protecting the cells within a system and working to keep them within their optimal temperature and voltage ranges in order to avoid a catastrophic failure that could cause harm to the people that are using the technology the lithium-ion cells are used in. Fig. 36 shows a simplified example of a safety map from a voltage and temperature perspective; the figure shows where there is a voltage range where the cell will operate without any concerns about safety in the green area. At the cell and chemistry level cell temperature and voltage are two directly measurable characteristics that we can use to determine the state of the battery. As that voltage limit is exceeded, either at high voltages or low voltages, this is where cell failure begins to occur. We’ll talk more about the types of failures that occur at these extremes shortly, but at high voltages we begin to see the cathode materials dissolving into the electrolyte which causes a domino effect of other failures to begin happening. Temperature effectively has a similar set of high and low limits, above and below those limits cell failure begins to occur. High temperatures have similar effect as operating at high voltages, but low temperatures tend to create lithium plating rather than lithium intercalation thereby reducing life and creating the potential for dendrites to form. Therefore looking at the chemistry and determining if there are things we can do to improve the performance at high and low temperatures, such as adding unique salts or additives to the electrolyte, or improving performance at high and low voltages, again possibly by using unique additives to reduce gassing at high voltages, is key to understanding cell and chemistry safety. The specific voltages and temperatures that will begin the failure mode differ depending on the chemistry and design of the cells. Some lithium-ion chemistries operate safely up to 4.2 volts while others max out about 3.0 volts, so trying to charge a lower voltage cell to the higher voltage limit will drive it into a failure mode

61

62

CHAPTER 3 Lithium-ion battery operation

in most cases. Generally, however, most lithium-ion chemistries operate within about the same temperature range, plus or minus a couple of degrees—from 20°C up to about +55°C. As we begin looking at lithium-ion cell failure modes we must start by asking the question what causes a cell to fail? In the end, cell failure comes down to three potential types of failure: internal short circuit, gas generation, or impedance growth. This is a bit of a simplification, as these are all outcomes from other events. For example, a short circuit could be internal, caused by debris generated during the manufacturing process or by the growth of dendrites that end up penetrating the separator and shorting the electrodes, or due to high temperatures melting the separator allowing the electrodes to touch. Or a short circuit could be the result of a foreign body penetrating the cell, like the infamous nail penetration test. But all these lead to a catastrophic failure event due to a short circuit. And pretty much every potential mode of failure ends up in one of these three types of failure. In Fig. 37 we see some of the possible causes of cell failures on the left side of the image, to the right of these we see the results that ultimately lead to cell failure and often thermal runaway event. We see here that either an internal or external short circuit, dissolution of active materials, lithium plating, and SEI growth can all lead to the three types of failures and in fact in many cases they will “stack” upon one another. These of course are meant to be simplified examples as there are a lot of other events that can occur to get the cell to these states. As one example let’s take a look at what happens at high temperatures, usually beginning above 90°C, where some interesting things begin to happen. For example, the traditional LiPF6 electrolyte may begin to break down above this temperature. Charged active electrode materials will react with the nonaqueous electrolytes at

FIG. 37 Cell failure types and causes.

3.9 Failures modes

high temperatures, often violently. As the temperatures rise the chemical breakdown of the cathode continues and the gases that are being generated begin to recombine into hydrogen and oxygen gases. This is one of the reasons lithium-ion cells limit the upper-end operating temperatures. In addition to the gases being generated here the high temperature will begin melting the separator which will cause a short circuit. So you can see both failure methods can happen during the same event. Yet even at room temperatures the lithium salts in the LiPF6 electrolyte have a tendency to react with moisture in the cell which results in the formation of hydrogen fluoride (HF) gas which will interact with the cathode active materials causing first surface erosion then reduction in capacity due to the degradation of the electrodes, electrolyte, and the growth of a new, thick passivation layer (Goriparti et al., 2014). This is also the reason that lithium-ion cell manufacturing must be done in a humidity-controlled environment and the reason that battery manufacturers include moisture limits on the incoming materials.

3.9.1 Internal short circuit There are several ways that a cell can end up with an internal short circuit beginning with the cell reaching high temperatures or due to the growth of a dendrite of lithium metal that penetrates the separator and allows anode and cathode to become electrically connected or even due to debris entering the cell during manufacturing (Fig. 38). The first manner a cell may experience a short circuit is during high temperature events. Most traditional separators are made of either polyethylene or polypropylene or a combination of the two. At temperatures as low as 90°C these separators’ pores

FIG. 38 Catastrophic short circuit events.

63

64

CHAPTER 3 Lithium-ion battery operation

begin closing and about 120°C–130°C the separator will melt and when that happens the polyethylene or polypropylene separators begin to shrink away from the electrodes. The shrinkage of the separator allows the anode and cathode electrodes to come into contact with one another, creating a short circuit within the cell and an internal short circuit. But what is really happening during an internal short circuit event in a lithium-ion cell? Briefly, most of the energy of the cell gets released very rapidly and through a single very small point. Some research has indicated that up to 70% of the energy of the cell may be released in less than a minute (Maleki & Howard, 2009). As most of the current in the cell is pushed through the internal short circuit site the rapid release of energy causes the cell to begin generating heat starting in the localized area around the internal short circuit. If the amount of heat that is generated is greater than the amount of heat that can be removed, the cell will continue to heat up until it reaches thermal runaway onset temperature. An important item to be aware of when it comes to any type of short circuit is the impact of the state of charge of the cell. Cells with a low state of charge will have less energy to release than will cells that are at a high state of charge. So there will be greater heating around the internal short circuit location if the cell is at a high state of charge compared to when it is at a low state of charge (Cai, Wang, Maleiki, Howard, & Lara-Curzio, 2011; Maleki & Howard, 2009). Another area to be aware of is the tendency for the growth of lithium “dendrites” during cycling. A dendrite is formed when the lithium metal begins plating on the anode. What occurs during dendrite growth is that there may be a spot near the negative electrode where there is a depletion of the salts in the electrolyte that may create a place where the chemical potential near the electrode surface is more positive (Huggins, 2009). This causes any bump or imperfection on the surface to become a home for lithium to plate and grow. If this continues it will grow into what appears like a stalagmite growing between the anode and cathode. If the dendrite continues to grow unchecked, it may penetrate through the separator causing direct electrical connection between the anode and the cathode (Fig. 39). This internal short circuit will force all the energy/power in the cell to discharge through that single small point generating heat, cell failure, and depending on the type of electrolyte flame. This failure mode is often seen during high rate charging. It is suspected that one of the causes of the Samsung Galaxy Note 7 fires in 2016 may have been due to dendritic lithium growth. This was not the sole cause of the cell failures but was likely one of the contributing factors that when they all occurred at the same time caused the cells to go into thermal runaway (Hruska, 2017a). The third manner that a cell may experience an internal short circuit is due to debris getting into the cell during the manufacturing process. One example of this happening was back in 2006 when Sony Corporation ended up recalling hundreds of millions of laptop batteries due to a high number of failures and thermal events. In this case the battery cell was a cylindrical 18650-type cell that used a nickel-plated steel can. During the assembly process the lid of the cell was crimped into place. It was found that this crimping process caused tiny particles of nickel to flake off inside

3.9 Failures modes

FIG. 39 Dendrite growth causing internal short circuit.

the cell causing an internal short of the cell (Wong, 2006). This was also one of the contributing factors in the Samsung recall mentioned before. Samsung’s investigation found that during the electrode welding process there were burrs that formed on the cathode which resulted in the burrs breaking through the separator and causing an internal short circuit event to occur (Hruska, 2017a, 2017b). This is one of the reasons that most battery manufacturers do their cell assembly in clean room/dry room environments, with both the humidity and particle size and count closely monitored. Clean rooms are rated according to how much particulate of specific sizes exist per cubic meter within the environment. By assembling the electrodes into the cells in a clean room environment the potential for foreign matter to find its way into a cell is greatly reduced. Finally, there are other types of failures that can get introduced to the cell during the manufacturing process. The Samsung Note 7 cells experienced several different failures including having the tops of the electrodes getting bent over during installation and assembly which caused the electrodes to touch, which in turn created an internal short. This was due to the manner in which the cell was installed in the phone where it created high stress on the cell in several areas and pushed the electrodes together. The other failures they experienced were due to an assembly failure where an insulating tape was not included during the assembly process and the electrode placements were misaligned causing internal stresses (Hruska, 2017a, 2017b). You can see from this example that it is not difficult to create a situation for a cell to create an internal short circuit. These types of manufacturing failures are difficult to identify ahead of time, but can be extremely costly both on the lives of those affected and financially on the companies responsible.

65

66

CHAPTER 3 Lithium-ion battery operation

3.9.2 External short circuit The other type of short circuit event that can occur is called an external short circuit. An external short circuit occurs when a conducting material that is external to the cell either touches the positive and negative terminals at the same time or is inserted into the cell causing an electrical connection with the electrodes to occur. The resulting failure mode of an external short circuit can be the same as an internal short circuit, rapid release of energy at single point, heat generation, gassing, spark, flame, explosion, and often ejection of the jellyroll from the case. In the short circuit event the duration of the event may initially be very rapid, with almost immediate heat generation and current rise in as little as 0.1 seconds. The next stage results in high current and maximum temperatures reached in as little as 120 seconds. From here the current and voltage slowly drop (Chen, Xiong, Tian, Shang, & Lu, 2016; Xia, Chen, & Chris Mi, 2014). If the heat generated during these stages is not removed the cell will continue into thermal runaway as described previously. The ability for a pack to safely manage one of these failure events is strongly dependent on the state of charge and the heat rejection capability of the pack. At higher state of charges, typically above 50%, the amount of energy released is enough to create high enough pressure in the can cell to allow it to begin venting electrolyte. The state of charge has a big influence on whether the cell is driven into thermal runaway or not. For example, if we take a typical NMC type cell as shown in Fig. 40 with a voltage range of 2.5–4.2 V and 15 Ah in capacity and look at it at 10%

FIG. 40 Comparing energy released at different SOC states.

3.9 Failures modes

SOC and at 90% SOC we see that there is a lot more energy being released, in fact about 37% more energy being released, when the cell is at the higher state of charge. For this reason the United Nations and the U.S. Department of Transportation (DOT) require that all lithium-ion cells and packs are discharged to not more than 30% state of charge before they are shipped.

3.9.3 Thermal runaway Thermal runaway occurs when a cell has reached the temperature at which the temperature will continue to increase on its own and it becomes self-sustaining as it creates oxygen which feeds the fire (literally). Once the temperature of the cell reaches about 80°C the SEI layer on the anode begins to decompose and break down in an exothermic reaction (generating heat) due to the reaction of the lithium with the solvents used in the electrolyte. At about 100°C–120°C the electrolyte begins to break down in another exothermic reaction, which in turn generates various gases within the cell. The gases that may be created during this reaction, depending on cell chemistry, include carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), ethane (C2H6), ethylene (C2H4), and hydrogen (H2) (Ohsaki et al., 2005; Wang et al., 2012). As the temperature nears 120°C–130°C the separator finally melts allowing the anode and cathode electrodes to make contact and cause an internal short circuit and generating more heat. As the temperature continues to rise, at about 130°C–150°C, the cathode begins breaking down in another exothermic chemical reaction with the electrolyte which also generates oxygen. It is this release of oxygen along with the carbonate LiPF6 electrolyte that ultimately allows the cell to burn and catch fire. The breakdown of the cathode active material is a highly exothermic reaction generating a lot of heat and continuing to drive the cell toward ultimate failure and fire. When temperatures rise above 150°C–180°C the reaction may become selfsustaining if the cell is not able to rapidly dissipate the heat being generated. At this point the cell is in what is referred to as “thermal runaway” as the oxygen generation makes the fire self-sustaining—at least until all the fuel has been used. If the gases continue to build up within the cell the cell may rupture or vent through a safety valve. The cell may rupture or vent the flammable hydrocarbon gases and hydrofluorocarbon electrolytes at this point and the introduction of a spark could ignite the electrolyte and the gases causing flame, fire, and potentially an explosion. But if the pressure continues to build up it is also possible that the cell will split open and eject the jellyroll from the housing (Wang et al., 2012). An example of a thermal runaway temperature map is shown in Fig. 41, but please note that these temperatures are not exact numbers because different chemistries and additives offer different performance. Through additives in the electrolyte, ceramic coating the separator and other tools the thermal runaway temperature can be raised significantly. In some cases, separators may be able to reach temperatures of 190°C or more without shrinking and causing the short that starts the thermal event. There are even some separators that appear to be able to continue to operate all the way up to 300°C and beyond, such as the Dreamweaver

67

68

CHAPTER 3 Lithium-ion battery operation

FIG. 41 Impact of temperature on a Li-ion cell.

family of separators. In other words, these temperature ranges are dependent on the cell design and chemistry. Some may move into thermal runaway at lower temperatures and others may move into thermal runaway at higher temperatures. During these events a variety of gases may be generated. The exact gases will depend on the specific electrolyte, cathode and anode formulations used. But for a traditional LiPF6 electrolyte with ethylene carbonate (EC) and ethyl methyl carbonate (EMC) salts the gases generated may include, in decreasing order, up to 50% carbon dioxide (CO2), up to 10% ethylene (C2H4), up to 10% hydrogen (H2), with lesser amounts of methane (CH4), ethane (C2H6), carbon monoxide (CO), tetracarbon (C4), ethyl fluoride (C2H5F), cyclopropane (C3H6), and propane (C3H8) (Roth & Orendorff, 2012). As a final note on thermal runaway, you will have recognized that I used the term exothermic at each stage of the process. This is important as an exothermic reaction is one that generates heat. The opposite of which is an endothermic reaction would absorb heat. It is the fact that all of these reactions are exothermic that generates heat during this process, each one adding to the heat generated by the previous reaction until it can no longer be contained.

3.9.4 Cascading failure One cell failing is a problem, but it can lead to a much larger problem if that cell failure drives the cells surrounding it to fail as well. This is referred to as a cascading failure event, when one cell goes into thermal runaway the heat generated may be great enough to drive the cells nearest it to also go into thermal runaway, this

3.9 Failures modes

cascades to the next cell which now adds to the heat being generated and then on to the next cell, and so on until the complete pack is engulfed. Think of it like a line of dominoes, once the first one falls it hits the second one and then continues until all the dominoes have fallen. A cascading failure in a lithium-ion battery pack will act in much the same manner if not controlled. The heat that is generated by a cell in thermal runaway has two major methods by which it can move to the surrounding cells, either through conduction or radiation. Conduction heating is heat transfer through direct contact. In the example image shown (Fig. 42), heat from the cell in thermal runaway is being transferred by conduction through the bus bar at the top of the cell that connects the two cells and to the cooling plate next to the cell. But the heat is also transferred by radiation to the cell next to it. The other method of heat transfer is convection which is done when heat is transferred through a moving liquid (air is considered a liquid in this instance). So as the air between the cells is heated up it can transfer that heat to other cells within the pack, especially if the pack uses an air-cooling methodology wherein the air is circulated within the pack (Warner, 2015). The challenge that is presented during a cascading thermal runaway event is in managing the heat that is being generated during these exothermic reactions. If the system can dispel or disperse the heat being generated quickly enough, then there is a chance of limiting the thermal runaway event to a single cell. However, depending on the size of the cell there can be more heat generated than is possible to remove before impacting other cells in the pack. In one case we studied the failure of a large format

FIG. 42 Methods of heat transfer.

69

70

CHAPTER 3 Lithium-ion battery operation

cell with a capacity of over 70 Ah and calculated the amount of heat generated as being more than a mega-joule of energy. That is roughly equivalent to about 950 BTUs of heat generation for a single cell, so the heat energy in the pack was equivalent to about one-third of a tank of gasoline. With the size of some of the largest lithium-ion applications reaching multiple megawatt hours in capacity you can understand the need to prevent cascading failures.

3.9.5 Impact and effects of temperature on cell aging Temperature has a major impact on performance and life of lithium-ion batteries. At low temperatures the ionic conductivity of the electrolyte gets significantly reduced making it more difficult for the ions to move from anode to cathode. The lower the temperature the lower the conductivity. As temperatures drop below 0°C both discharging and charging become more difficult. At 10°C charging typically must be halted and discharging rates need to be reduced. At 20°C charging may need to be discontinued and discharging may not be possible as the ionic conductivity in most LiPF6-based electrolytes comes down to near zero. At low temperature there is another factor that comes into effect, lithium plating. As the temperature drops the ability of the lithium to properly intercalate into the anode during charging is limited which causes the lithium to plate as lithium metal on the surface of the electrode. This has an immediate effect of reducing the capacity of the cell by reducing the amount of lithium for future intercalation as the plated lithium is no longer mobile and cannot participate in the charge/discharge process. In the long term, the plated lithium may form dendrites which we have already discussed may result in a potential internal short circuit. At high temperatures the separator begins to break down, the electrolyte begins chemically reacting with the cathode creating the forerunners for the thermal runaway events as described earlier. Most standard polyethylene or polypropylene separators begin shrinking at between 90°C and 110°C as was already discussed. This will cause an internal short circuit within the cell and gas production at the interface between the cathode and the electrolyte as the temperature continues to rise. Temperatures above 120°C–130°C cause the dissolution of the cathode into the electrolyte. Once the activity of the dissolved gas exceeds critical solubility levels, gas bubbles are formed (Lefrou et al., 2012, p. 65). This in turn may lead to thermal runaway, fire, and even potential explosion. But even looking at operating a cell at temperatures above the recommended average but below the safety limit will also have negative effects on the cell. Operating a cell regularly at 45°C or 50°C will have the same effects due to cathode dissolution and gassing as noted before, but just at a much slower pace. Operating your cell at these temperatures will cause your cycle life to drop off rapidly. Now temperature can be a bit of a tricky thing because some chemistries, such as many NMCbased cells, will see an increase in capacity up to about 35°C. But this is a short-term effect and will ultimately cause premature failure of the cells if they are operated at this temperature range for too long.

3.9 Failures modes

3.9.6 Impedance During the life of a lithium-ion cell, the cell will slowly increase in internal impedance. Impedance is effectively the resistance to the movement of ions, electrons, and current in a cell. The higher the impedance of a cell the harder it is to push lithiumions through it. This is either from operation at extreme temperatures, either high or low, or as a result of the growth of the SEI layer over time or a combination of these. The impedance of a lithium-ion cell has three parts. The first part is the sum of the ionic resistance of the electrolyte. The second is the electrical resistance of the active materials, such as the current collectors and foils. The third is the contact resistance of the active material and the current collector. The sum of these three types of resistance is the internal impedance of a cell (Dahn & Ehrlich, 2011; Salomon, 2011). Impedance can significantly influence the operation of a cell as it will cause a drop in the voltage as moving lithium-ions in a high impedance cell consumes some of the energy in the form of heat. At low temperatures impedance also increases significantly due to the reduced ionic conductivity of the electrolyte, usually at temperatures below 20°C. Another impact of increased impedance is that over time the SEI layer of most lithium-ion cells will increase in thickness, in some cases by as much as 100% or more. This is generally a result of additional lithium plating at the SEI layer over time which reduces the available surface area for the chemical RedOx reactions to occur as well as consuming some of the free lithium all of which reduces available capacity. As current is continuously moved through the cell this SEI growth slowly reduces the life of the cell. A good example of this is the battery in your tablet computer or smart phone. These types of devices rarely experience temperature variations, nor do they see high power pulses that could drive this impedance growth. Yet over time you begin to notice that the amount of time your battery can operate your laptop or cell phone without being charged gradually get smaller and smaller. This is the result of this impedance growth over time. While an increase in impedance in a cell will have an impact on cell life, it is generally not a driver of catastrophic cell failures.

3.9.7 What happens during overcharge A cell can only be overcharged when it is connected to a power supply. Remember that the natural state of a battery is to discharge so it will naturally lose capacity without an external power source, not gain it. The main reasons a cell (or battery) may be overcharged are either due to a battery management system that is not designed properly or a malfunction of the charging system. However, one other overcharge mechanism may be mentioned here. In a system that includes many cells in parallel and series it is possible that a single cell may have greater, or smaller, internal resistance due to manufacturing variation, location of the cell in the pack, or many other possible causes. This may cause that one cell to be overcharged more than the others around it which could eventually lead it into a failure and thermal runaway event if the overcharging is severe enough (Ye et al., 2016).

71

72

CHAPTER 3 Lithium-ion battery operation

The voltage at which a cell will move into overcharge will differ depending on the chemistry of the cell. For NMC, NCA, LCO, and LMO-type chemistries the top operating voltage is about 4.2 volts. LFP cells have a maximum operating voltage of about 3.6 volts and LTO chemistries have an upper operating voltage of about 2.8 volts. Overcharge begins as those chemistries begin to exceed those upper operating voltage limits. The impact of overcharging a cell will depend on how much it is overcharged and for how long. Most chemistries can accept small amounts of occasional overcharging with minimal impact on the performance or life of the cell. But continuous overcharging will result in a series of failures that can lead to a thermal runaway event. In the first stage of a cell overcharge the voltage and temperature rise slowly as the lithium-ions are permanently forced from the cathode and into the anode, this is an irreversible reaction. In the next stage, temperature and voltage continue rising but much more quickly. The voltage will hit a high peak and then drop off precipitously as the cathode is fully delithiated. During this stage the electrolyte begins to react with the cathode. At this point the electrolyte begins to decompose and the metals in the cathode, such as nickel, manganese, cobalt, and iron, may begin to get dissolved from the cathode material and begin plating on the separator and anode. This causes an increase in internal resistance in the cell by as much as 500%. This increased impedance also causes an increase in the ohmic heating inside the cell which in turn causes the continued dissolution of the cathode and decomposition of the electrolyte and the generation of gases within the cell (Ohsaki et al., 2005; Yuan et al., 2015). As the gases continue to evolve and the temperatures continue to rise the cell will begin to move into a thermal runaway event as the temperatures begin to exceed about 130°C. As temperatures reach 180°C, which is the melting point of lithium, the thermal runaway becomes irreversible. Thermal runaway may be due to a violent reaction between the overcharged anode and the high temperatures of the electrolyte that result in an exothermic reaction between the delithiated cathode and the electrolyte (Ohsaki et al., 2005). At this point the cell is likely to rupture to release the gases and may expel the jellyroll potentially with flame and explosion.

3.9.8 What happens during overdischarge While overcharging a lithium-ion battery is more common than overdischarging the battery, significantly overdischarging a lithium-ion battery has similar undesirable effects on the performance, life, and safety. There are four main methods that a cell (or battery) may be overdischarged. First, a cell (or battery) may be forced to continue discharging due to the controls being inadequate or not having the correct end voltage. Second, a cell (or battery) may overdischarge due to a quiescent current, in other words due to the continuous power drawn from the electronics in the system even when the system is “powered down”. A third method for an overdischarge event could be due to an internal short circuit in the cell, this may start as a filament or

3.9 Failures modes

dendrite growth or it could be due to a manufacturing flaw that introduced a foreign element into the electrodes. The fourth method that a cell (or battery) may be overdischarged is due to the self-discharge that can occur during long-term storage (Maleki & Howard, 2006). Of course, as we often say in the battery world the end result depends on many factors. For a typical NMC cell the operating voltage range is about 2.5 volts at the bottom to 4.2 volts on the top end. Discharging the cells down to 2.5 volts will generally not have any impact on the cell performance. When we are talking about overdischarging of a cell, we are really looking at discharging down below that minimum operating voltage level. In the example of an NMC cell, this means discharging it down to 2.0, 1.5, 1.0 volts all the way down to 0 volts. For LFP chemistries which have an operating range of about 2.5–3.6 volts, overdischarging would begin as the cell is discharged below the 2.5 volts minimum operating voltage. With an LTO-based chemistry the operating voltage is a bit lower still, operating from 1.5 to 2.8 volts, so in this case overdischarging occurs as the cell is discharged below 1.5 volts. Overdischarging a lithium-ion cell will result in a permanent reduction in capacity as well as the potential for internal short circuit as described before. During the overdischarging event the voltage potential of the anode increases which causes the copper current collector to begin to oxidize. At the same time the lithium-ions that were inserted in the anode begin to deintercalate and move into the cathode. As the overdischarging process continues the cathode is affected as its base morphology begins to get changed because of these various “side reaction” that are occurring, including the deintercalation, SEI breakdown, and copper film oxidation. This over deintercalation causes the SEI layer to begin to decompose, which generates gases. When the cell is again recharged a new SEI layer gets formed on the anode surface. These factors cause a permanent reduction of the cell’s capacity (Guo, Lu, Ouyang, & Feng, 2016; Maleki & Howard, 2006). The other impact of overdischarging a cell is the risk of reversing the polarity of the cell. This occurs when you attempt to continue to push current through a cell that is fully discharged. This will cause the polarity of the cell to swap, positive becomes negative and negative becomes positive. In a large system this may have caused this cell to be overtaxed and ultimately drive it into a thermal failure event at worst and at best will significantly reduce the life of the pack as the rest of the cells will have to work much harder to overcome the reversed cell. In addition to the impacts on life, there are some significant impacts on safety as the lithium-ion cell is overdischarged. As mentioned before one of the side reactions that occur is the oxidation and dissolution of the copper current collector. The copper ions that get dissolved into the electrolyte will flow through the separator and begin accumulating on the cathode. As this accumulation increases it begins to form copper dendrites. If these dendrites continue to grow they will penetrate the separators and cause an internal short circuit of the lithium-ion cell (Guo et al., 2016; Maleki & Howard, 2006).

73

74

CHAPTER 3 Lithium-ion battery operation

3.9.9 Influence of impact, crush, and penetration When it comes to physical damage and abuse of lithium-ion cells, the failure mechanisms are those that we have already covered including internal short circuit, external short circuit, and thermal runaway. As I have already said, different cells with different chemistries perform differently. Some cells can withstand an abusive condition such as a penetration without driving it into thermal runaway, fire, and explosion while other chemistries will almost immediately move into a thermal runaway event. During a nail penetration several things occur virtually simultaneously. First, as the nail penetrates the enclosure of the cell and then the electrodes it will cause fracturing of the packaging pulling pieces of enclosure and electrodes into contact with each other causing a short circuit. At this point all the current in the cell is released causing heat, fire, flame, and explosion in many cases. Impact and crush testing either drops a device onto the cell from between 1 to 3 meters in distance or forces a pressure on the cell until it reaches about 25% of its original thickness or thermal runaway onset occurs whichever occurs first (Zhang, Ramadass, & Fang, 2014). In many cases these types of failure modes are being mitigated at the module and pack level. In one study conducted by the MIT Crash Laboratory they evaluated the influence of different types of road debris penetrating the pack enclosure and the impact it had on the cells within the pack. They ultimately recommended using a steel plate to act like a piece of armor on the bottom of the pack to protect the pack from road debris penetrating the pack. This may work to protect the pack but adds significant weight to the pack as well as some additional cost (Xia, Wierzbicki, Sahraei, & Zhang, 2014).

3.9.10 Aging mechanisms The final failure mode we will review is also a question that is almost always asked at some point in the initial discussions, how long will the lithium-ion battery last? But that is somewhat of a trick question as there are a lot of different factors that determine the life of a battery. When it comes to lithium-ion battery aging there are two major types of aging to be considered. The first is based on evaluating the number of charge and discharge cycles (cycle life) that can be achieved under a variety of depth of discharge and temperature conditions. This one is the most variable as it is entirely dependent on the actual charge/discharge profile that it is being tested under. The second is based on calendar life and measures the capacity of a cell at the beginning and then lets it sit on a shelf at a consistent temperature for months or even years and then measures the capacity at the end both before and after cycling. Some amount of energy will be permanently lost during this testing and some will be able to be recovered through cycling the battery. Cycle life is perhaps the more important of the two, but it is also quite variable in that the cycle life achieved will depend on the application, the temperature, how fast it is charged, how fast it is discharged, and the depth of discharge among many other

3.9 Failures modes

factors. For instance, an all-electric transit bus that charges once a day overnight but operates 7 days a week and 365 days a year for 10 years, it will need to achieve 3650 cycles to reach that 10-year life (365 cycles per year  10 years). In addition, if the battery only uses 80% of the available energy to achieve that 3650 cycles then its 100% cycle life rate might only need to be 1500 or 2000 cycles. But what if the first bus went through the cycles faster? If the operators decided to charge it once during the day and once at night that would increase the cycle usage so now the same bus may only reach a 5-year life using the same set of parameters instead of the 10-year life expected. It will also change if the bus is being operated in Phoenix, Arizona where the ambient temperature is much higher for longer periods of time than if it is operating in Milwaukee, Minnesota. Since temperature has such a strong impact on cycle life, understanding where the application will be used will directly impact the cycle life estimates. This is a relatively common challenge when sizing a battery for an application that may have to operate over a very long time span. Since the way the application is used is likely to change between the first day of its life and the last day of its life. This means that it could face a harsher operating cycle or an easier one at various points during the battery’s life, with either having an impact on overall cycle life. This is one of the reasons that you frequently hear battery professionals answer with “it depends” when asked how long a battery will last! As another example, imagine another all-electric transit bus that wants to fast charge its batteries once an hour for 10 hours or 10times per day. Over the same 10-year life that battery would require 36,500 charge/discharge cycles (10 charges per day  365 days per year  10 years). This battery may need to reduce the amount of usable energy down to 60% DOD to achieve that cycle life, which means that the 100% cycle life may need to be 15,000–18,000 cycles which will also eliminate several chemistries as potentials for this application. Temperature also has a significant impact on the overall life of a lithium-ion battery. As mentioned in my first book (Warner, 2015) batteries are comfortable about the same temperatures that you are. When temperatures drop or rise outside of that comfort zone, the performance of the batteries drop and with repeated exposure will dramatically reduce the life of the battery. Another aspect of cycle life is the power usage during the cycle. A bus operating in a very hilly area would require more power capability than a similar bus operating in a flat route. All these factors influence the life of the battery. But with all that said, what is really happening inside the cell that causes the aging of the cell? While there are a couple of different factors that impact lithium-ion cycle life, the biggest impact is due to the SEI layer. Over time the continuous insertion and deinsertion of the lithium-ions into the graphite layers during cycling may cause the SEI layer to be stretched and potentially destroyed. Once this occurs a new chemical reaction between the electrolyte and the graphite anode will cause a new SEI to be created to repair or replace the damaged area. However, the chemical reaction that creates this new SEI area uses up some of the free lithium in the electrolyte, thereby permanently reducing the overall capacity of the cell. This corrosion of the anode and

75

76

CHAPTER 3 Lithium-ion battery operation

decomposition of the electrolyte continue very slowly over the life of the battery causing the SEI to slowly penetrate into the pores of the anode and even into the separator which results in a reduction in the amount of accessible surface area of the anode, increasing internal resistance, and causing a permanent loss of capacity and power (Barre et al., 2013; Leng, Tan, & Pecht, 2015; Troltzsch, Kanoun, & Trankler, 2006; Vetter et al., 2005). Other aging mechanisms include structural changes to the cathode and anode, the decomposition of the electrolyte, the dissolution of the active material into the electrolyte, and the creation of a SEI-type layer over the electrode surfaces and the current collector surface. On the cathode side of the cell the active materials begin to decompose at voltages above approximately 4.35 volts causing an increase in charge transfer resistance within the cell. At high temperatures the cathode also begins to experience structural and phase changes which reduce the ability to intercalate the lithium-ions during cycling. The structural changes may drive a change in the crystalline structure of the cathode, changing it into a cubic phase or a spinel phase structure both of which reduce the charge transfer rate (Leng et al., 2015). One other impact of the cathode dissolution is that, depending on the chemistry, it may release transition metal (cathode) ions that may get incorporated into the SEI thus reducing the capacity and increasing the thickness of the SEI layer (Vetter et al., 2005). Another mechanism that causes lithium-ion battery aging is due to repeated overcharging and overdischarging the battery. This leads to the decomposition of the electrolyte, reducing ionic conductivity and increasing the internal resistance of the cell. The current collectors may also begin to corrode over time or even begin to be dissolved during operation. This increases the internal resistance of the cell over time (Troltzsch et al., 2006). It is also very important to recognize that all of these aging mechanisms are greatly accelerated at higher temperatures causing rapid deterioration of the cell performance. At high temperatures the active material of the electrodes begins to decompose rapidly. Some studies have found that another SEI layer gets created at the cathode interface with the electrolyte at high temperatures thereby increasing the resistance of the cell and leading to poor reintercalation of the lithium-ions in the cathode. And at elevated temperatures the SEI on the anode may see a replacement of the carbon with an inorganic species in its place, increasing resistance and increasing the thickness of the SEI layer (Leng et al., 2015). At low temperatures the cell may experience lithium metal plating on the SEI layer or lithium dendritic growth on the SEI—both of which reduce the capacity and power due to the permanent loss of the lithium-ions and accelerate the aging of the cells while creating the potential for future safety issues (Vetter et al., 2005). All of these different aging mechanisms are summarized in Table 2 (Vetter et al., 2005) along with their causal factors and the impacts.

3.9 Failures modes

Table 2 Aging mechanisms of lithium-ion batteries Accelerated by

Cause

Effect

Leads to

Reduced by

Electrolyte decomposition (to SEI) (continuous side reaction at a low rate) Solvent cointercalation, gas evolution, and subsequent cracking formation in particles Decrease of accessible surface area due to continuous SEI growth Changes in porosity due to volume changes, SEI formation, and growth Contact loss of active material particles due to volume changes during cycling Decomposition of binder

Loss of lithium, impedance rise

Capacity fade

Stable SEI (additives), rate decreases with time

High temperatures, high SOC (low potential)

Loss of active material (graphite exfoliation), loss of lithium

Capacity fade

Stable SEI (additives), carbon pretreatment

Overcharge

Impedance rise

Power fade

Stable SEI (additives)

High temperatures, high SOC (low potential)

Impedance rise, overpotentials

Power fade

External pressure, stable SEI (additives)

High cycling rate, high SOC (low potential)

Loss of active material

Capacity fade

External pressure

High cycling rate, high DOD

Loss of lithium, loss of mechanical stability Overpotentials, impedance rise, inhomogeneous distribution of current and potentials Loss of lithium, loss of electrolyte

Capacity fade

Proper binder selection

High SOC (low potential), high temperatures

Power fade, enhances other aging mechanisms

Current collector pretreatment

Overdischarge, low SOC (high potential)

Capacity fade, power fade

Narrow potential window

Low temperature, high cycling rates, poor cell balance, geometric misfits

Current collector corrosion

Metallic lithium plating and subsequent electrolyte decomposition by metallic lithium

77