The metal-air battery developed by IBM over a seven-year period has a range of 800 kilometers.

    IBM started researching metal-air batteries seven years ago. How well do they perform? What are the obstacles to mass production?

    Many people may not realize that IBM has been researching battery technology for a long time. In 2009, IBM started a project called Battery 500 at the IBM Almaden Research Center in San Jose, California, with the goal of developing a battery that can extend the range of electric vehicles to 800 kilometers.

    The two authors of this article, Winfried W. Wilcke and Ho-Cheol Kim, are both from IBM Almaden Research Center. The former is the head of nanotechnology research and the latter is the head of the energy storage research group of the research center.

    The original article was published on the IEEE website and translated by Che Yunjun. The two authors jointly described the advantages and disadvantages of metal-air batteries, the various problems encountered on the road to commercialization, and how they were solved.

   Why is 800 kilometers the goal? Because this value is the maximum value that most people expect for the range of a car. If the range of an electric car cannot reach 800 kilometers and the cost is acceptable to most people, then electric cars will have less chance of becoming popular.

    So, we set this value as the goal of our Battery 500 project. This project started in 2009 and was led by the Almaden Research Center. Since then, IBM has conducted this research together with many business partners and research institutes from Europe, Asia and the United States.

    The Battery 500 project is based on metal-air technology. Compared with lithium batteries, metal-air batteries can have more energy per unit mass. The project is still a few years away from commercialization. However, through the seven years of experiments, we can conclude that metal-air batteries will be useful in electric vehicles in the future.

Why is it called a metal-air battery?

    Taking lithium-air batteries as an example, to understand this question, we first need to look at the difference between lithium-ion batteries (i.e. the common lithium batteries today) and lithium-air batteries.

    The following figure shows the internal state of a lithium-ion battery during charging and discharging. In a traditional lithium-ion battery, the positive electrode is carbon, while the negative electrode is composed of different transition metal oxides, such as cobalt, nickel, manganese, etc. Both electrodes are immersed in an electrolyte containing lithium salts. During charging and discharging, lithium ions move from one electrode to the other. The direction of movement varies depending on the battery state, charging or discharging. During charging and discharging, lithium ions will eventually embed into the atomic layer of the electrode material, so the final capacity of the battery depends on how much material can accommodate lithium ions, that is, it is determined by the volume and mass of the electrode.

△Schematic diagram of the charging and discharging process of lithium-ion batteries

    Lithium-air batteries are different. In metal-air batteries, an electrochemical reaction occurs. During discharge, the positive electrode containing lithium releases lithium ions, which move to the negative electrode and react with oxygen on the surface of the negative electrode to form lithium peroxide (Li2O2).

    Lithium ions, electrons and oxygen react on the surface of the negative electrode formed by porous carbon. Because the chemical reaction does not occur on the negative electrode, it is not the negative electrode material that ultimately holds the lithium ions. Therefore, the capacity of the battery has little to do with the volume or mass of the negative electrode material, as long as there is a large enough surface area.

    That is to say, the capacity of lithium-air batteries is not determined by the volume and mass of the electrodes, but by the surface area of ​​the electrodes. This is why in lithium-air batteries, electrodes with very small mass can also store a large amount of energy, thus obtaining a higher energy density.

 △Schematic diagram of the charging and discharging process of lithium-air battery

    Of course, in addition to energy density, cost is also a very important consideration. The price of batteries is currently $200-300 per kWh. If each kWh can run 5-6 kilometers, a 150 kWh battery is needed for 800 kilometers, which will cost $30,000-45,000. A BMW 2 Series car only costs $33,000. Therefore, if you want to mass produce, the price per kWh must drop below $100.

What problems must be solved for the commercialization of lithium-air batteries?

    When lithium and oxygen undergo redox reactions, the theoretical maximum energy density that can be generated is 3460Wh/kg. Leaving aside the part of the battery cell that does not undergo chemical reactions, the energy density that can be achieved is also very exciting. Of course, there will also be problems.

    The charging process of lithium-air batteries is similar to that of conventional lithium-ion batteries, which can be achieved by applying external pressure. The difference is that in lithium-air batteries, when there is external voltage, the structure of lithium peroxide will be destroyed and reduced to oxygen and lithium ions, and the lithium ions will return to the positive electrode. Like traditional lithium batteries, the more times lithium-air batteries are charged and discharged, the greater the side effects generated inside the battery. These side effects are the fundamental factor affecting their mass production and even commercialization.

    To understand the impact of these side effects on the battery, we used the center's electrochemical mass spectrometer to accurately measure the amount of gas consumed and produced during each charge and discharge cycle. The result was a problem: lithium-air batteries release much less oxygen during charging than they consume during discharge. (In most experiments, dry oxygen is used instead of air.)

△Electrochemical mass spectrometer at IBM Research Center (Source: IBM)

    In an ideal battery cell, the mass of oxygen consumed during discharge is equal to the mass of oxygen released during charging. However, the study found that the amount of oxygen released was reduced, which means that the oxygen that was not released was likely to react with the components in the battery cell, such as melting into the electrolyte, and the battery was consumed internally.

    In another IBM lab in Zurich, we conducted new experiments to track and simulate this self-destructive chemical reaction. We finally found that the cause was the organic electrolyte. Then we studied this problem and used a new electrolyte in the latest battery cell. When charging, it can release most of the oxygen absorbed during discharge. In addition, we also tracked the consumption and production of hydrogen and water during charging and discharging, because the presence of these two substances means that there is a high probability that there is at least one self-destructive chemical reaction inside this battery. Our current battery cell can reach 200 charge and discharge cycles, although this makes the actual charging process far less than the theoretical maximum.

    Beyond this question, we made some key discoveries about the various components of lithium-air batteries:

    1. Positive electrode

    Unlike the positive electrode made of graphite in traditional lithium-ion batteries, the positive electrode containing lithium in lithium-air batteries undergoes some changes on the surface during charging, growing some moss-like or tree-like structures called dendrites. These dendrites are very dangerous because they can form a conductive loop between the positive and negative electrodes, causing a short circuit.

   △The positive electrode of the lithium-air battery has dendritic structures on its surface after dozens of cycles.

    In order to reduce the generation of dendrites, we use a special separator. This separator is composed of a layer of material containing many nano-scale pores. These pores are small enough and evenly distributed on the membrane to allow lithium ions to pass through and suppress the generation of dendrites. Because of the existence of this separator, the positive electrode can maintain a smooth surface after hundreds of charging cycles. If a traditional separator is used, dendrites will be generated after a few cycles. If a glass polymer containing conductive ions is used, the effect will be better.

   △The positive electrode of lithium-air battery, after using nano-isolation membrane, the surface remains smooth

    2. Electrolyte

    The electrolytes currently used still react with oxygen or other compounds produced during the charge and discharge cycle and are consumed. So far, no solvent has been found that is stable enough to allow lithium-air batteries to enter the commercial stage.

    3. Cathode

    During the charging process, lithium ions may react with the negative electrode to produce lithium nitrate. Lithium nitrate will also react with the electrolyte, consume the electrolyte and produce carbon dioxide. In our experiment, we also tracked the amount of lithium nitrate produced and took some measures to reduce its production. However, because the external charging voltage must be at least 700mV higher than the operating voltage of the battery. Overvoltage will reduce the charging efficiency of the battery. We have tried replacing carbon with some other metal oxides, and the results did not change much.

    4. Catalyst

    There have been many debates on whether to use catalysts in metal-air batteries. The use of catalysts can significantly reduce the occurrence of overvoltage conditions, but catalysts also usually accelerate the consumption of electrolyte. In our theoretical studies, the activation energy of lithium oxidation and reduction reactions is very low, so catalysts are not necessary in lithium-air batteries.

    5. Preparation of air

    Although the battery is called a lithium-air battery, we actually use dry oxygen. The emphasis on "dry" is because we only need to remove the water vapor and carbon dioxide components in the air. To produce such air in large quantities in commercial batteries, we need a sufficiently light, efficient and stable air purification system. From this perspective, the earliest practical application of lithium-air batteries may be in buses, trucks and other large vehicles, and only these large vehicles can accommodate air purification equipment.

    The size of the battery cells used for testing is still very small, with a diameter of 76mm and a length of 13mm, which is far from the standard for electric vehicles. Therefore, an important task that needs to be done is how to make larger battery cells and package many battery cells into a battery pack, and then equip it with a battery management system. We are also testing some different sizes, such as 100×100mm (100mm diameter, 100mm length).

    At present, this project is still in the initial basic science stage of materials and chemical reactions, but fortunately the research results are positive. In our research, the energy density that can be achieved now is 15KWh/kg for lithium redox reaction (using the original carbon cathode, 5700mAh×2.7V/g), and the energy density in the battery cell is about 800Wh/kg.

Sodium-air battery: low energy density, but stable

    There are many metals that can be used in metal-air batteries, including sodium and potassium in addition to lithium. The reverse reaction of these metals is easier, and relatively heavier metals, such as magnesium, aluminum, zinc, and iron, have been proven to be difficult to recharge, so the Battery 500 project finally chose to study lithium and sodium.

    Sodium-air batteries are another interesting combination. Although they may achieve lower energy density than lithium-air batteries, they have the advantage of being more stable.

    The reason for the lower energy density is that the chemical reactions are different. As mentioned earlier, in lithium-air batteries, lithium reacts with oxygen to produce lithium peroxide (Li2O2), but in sodium-air batteries, sodium reacts with oxygen using only one electron, producing sodium superoxide NaO2 instead of sodium peroxide Na2O2. In comparison, the energy density that sodium-air batteries can produce is theoretically reduced by half, and the theoretical upper limit of energy density is 1100wh/kg.

    But on the other hand, the charging efficiency of sodium-air batteries is higher than that of lithium-air batteries, and the overvoltage is quite low, less than 20mV (700mV for lithium). In view of this, the operating voltage of the battery cell can be reduced to 3V, so that the self-consumption of other components inside the battery can be greatly reduced, such as the electrolyte. We measured it through experiments and verified it. The advantage of this is that the stability of the battery is quite high. After 50 charge and discharge cycles, the capacity of the battery has hardly changed.

    There are also some challenges in the commercialization of sodium-air batteries. For example, sodium-air batteries consume twice as much oxygen as lithium-air batteries during the reaction, which is equivalent to the amount of air required for a piston engine that can produce the same power. In addition, the chemical activity of sodium metal is quite high. I believe many people remember the demonstration made by the chemistry teacher in high school class. A small piece of sodium thrown into water will cause a violent chemical reaction.

    However, lithium is a rare metal and is not cheap. But sodium is a common metal and has a very low cost. The cost of materials in a sodium-air battery of the same size is less than one-tenth of that in a lithium-air battery. Although lithium-air batteries will have better performance in the long run, considering stability and cost, sodium-air batteries, which have a high specific energy, will be a better choice from current batteries to the future.