By Keith Button|March 2023
Lithium-air batteries have intrigued futurists with their promise of storing vastly more electricity than today’s lithium-ion versions. But they have always suffered from an Achilles’ heel: They couldn't be charged and discharged over and over again, as required for commercial applications, including air travel. Keith Button spoke to researchers who have made a breakthrough in that part of the lithium-air equation.
On the campus of the Illinois Institute of Technology, researchers strode into Perlstein Hall daily to check on their baby — a black and gray disk the size of two stacked dimes, with wires leading to a computer whose display showed voltage measurements. This one-cell battery was their attempt to push research ahead on lithium-air batteries, a still experimental class that has tantalized battery researchers and aviation futurists for at least a decade.
What’s the attraction? A lithium-air battery draws in ambient air and culls oxygen molecules from it. This way, there is no need for the metal strips that in lithium-ion batteries feed the oxidization reactions that store electricity during charging. Likewise, there is no need for the graphite lattice structure that holds lithium and supplies lithium ions so electricity can be drawn from the battery. Most important, though, might be the direct lithium-oxygen chemical reactions that lead to storing more electricity than is possible with the metal oxide reactions in lithium-ion versions. In fact, by one estimate, lithium-air batteries could someday store five times more electricity per kilogram than a Tesla battery, giving these batteries a potentially far superior specific energy.
Scale up a single cell 10 to 15 times and put thousands of those together to form a large battery, and several of these batteries could, optimistically speaking, power a 100-passenger regional plane. So why aren’t today’s aircraft flying around on electricity supplied by lithium-air batteries? The Achilles’ heel has been the tendency of these batteries to wear out too soon when subjected to multiple cycles of charging and discharging. Researcher Larry Curtiss of the nearby Argonne National Laboratory set an informal benchmark of 1,000 cycles as the threshold for commercial viability, and by extension, that would include aviation.
At Perlstein Hall in September 2021, researchers led by Mohammad Asadi, an assistant professor of chemical engineering, were in the midst of testing their solution to the problem. At the time, no one had made a comparable lithium-air battery last through more than 200 cycles. They planned to run their battery to failure to see how many cycles it would survive. The answer turned out to be slightly beyond the threshold of 1,000 cycles.
Afterward, Asadi and his team of then-doctoral students Alireza Kondori and Mohammadreza Esmaeilirad and postdoctoral researcher Ahmad Mosen Harzandi enlisted other specialists to analyze the battery’s chemistry. This larger team, 17 in all, published its findings in the Feb. 3 issue of Science: “A room temperature rechargeable Li2O-based lithium-air battery enabled by a solid electrolyte.”
How did Asadi and his colleagues make the single-cell battery behind their rechargeability breakthrough? They started with one key constraint: To be practical, the battery had to operate at room temperature. Previous room temperature designs contained a liquid electrolyte, the material that carries ions between the cathode and anode. But in 2019, Asadi set out to create a solid-state version, meaning one with a solid electrolyte. He wasn’t sure how rechargeable the battery would be, but at a minimum, he saw two potential advantages: safety, and the ability to reduce weight and volume.
Liquid electrolytes are typically flammable, which can create a bad combination in a battery. “You have heat; you have oxygen; you have flammable materials. That makes for fire,” he says.
A solid-state battery — one with a solid electrolyte — would be lighter and easier to manufacture than a battery with a liquid electrolyte that would need to be contained. The solid electrolyte also would be somewhat pliable, so it could be shaped into any geometry that is convenient.
“It gives you more freedom and flexibility for scaled-up device applications in different variant architectures,” Asadi says.
The question was what material should serve as the solid electrolyte. Today’s generation of lithium-ion batteries powering our cars and phones contain liquid electrolytes, but Asadi knew that researchers were developing solid-state versions. He and his colleagues looked at these experimental lithium-ion batteries and realized they would need an alternative to the unsatisfactory choices they saw. One of those choices was a ceramic that contained lithium. This material was highly conductive, but it also would make a poor electrical interface with the anode and cathode. The other choice was polyethylene oxide, a polymer that lithium ions move easily into from the cathode or anode, and easily back out of. The downside was that ions would not pass easily through this polymer once inside it.
They decided to mix the two in a manner that would combine the best traits of both: An electrolyte that would allow ions to pass through easily and also easily cross the interfaces with the anode and cathode. They mixed up batches of alternatives, tested the electrical properties and tweaked the ingredients for the next batch, repeating the process until they settled on their best candidate. The winner consisted of particles of ceramic measuring just 70 nanometers in diameter — less than half the size of a grain of flour. They coated these grains with a polymer and then mixed them into the polyethylene oxide. The polymer coating on each particle provided a strong chemical bond for ions to pass between the ceramic particles and the polyethylene oxide.
They mixed up the electrolyte as a slurry, poured it over a petri dish, let it dry for three days at room temperature, then dried it for two more days at 50 degrees Celsius (122 Fahrenheit) under a vacuum.
Following the testing in 2021, Asadi looked at the results of the ionic conductivity testing of the electrolyte, meaning how well it allowed lithium ions to pass through. “This is very good,” he remembers thinking. Also impressive was the specific energy. The one-cell battery stored 685 watt-hours per kilogram, and the researchers estimated that with design tweaks, they could have reached at least 1,000 Wh/kg. That’s about three times the capacity of the most advanced lithium-ion battery, but still somewhat short of the projected ceiling of beating a Tesla battery by five times.
But it was cracking the 1,000-cycle threshold that was most exciting. At this point, the Asadi team did not know precisely what factors within their cell produced that result. So Asadi arranged a Zoom call with his mentor, Curtiss, who had been researching lithium-air technology for about a decade at Argonne, the U.S. Energy Department-funded research lab.
“The results were really, really amazing,” Curtiss says.
Curtiss and Asadi rounded up the team of 17 to find out what went on inside the cell. They examined the cathode and electrolyte with various methods. Specifically, they were looking for chemical reactions that would explain the cell’s ability to cycle repeatedly and with a large storage capacity. They knew that in a lithium-air battery, the key energy-producing reaction occurs on the cathode. As the battery discharges electricity, lithium ions flow from the lithium anode through the electrolyte to the cathode. There, the ions react with oxygen from the air to form one of three possible compounds: lithium superoxide, which requires one lithium electron per oxygen molecule; lithium peroxide, which requires two electrons; or lithia, which requires four electrons. These four-electron reactions would provide the most storage capacity, but historically, only a high-temperature lithium-air design incorporating a molten electrolyte had achieved such reactions. The analysis showed that Asadi’s cell created these four-electron reactions during its discharging and recharging cycles, the first time a room-temperature cell had done so. Liquid-electrolyte versions had created only one- or two-electron reactions.
“That’s why we think it’s a big breakthrough,” Curtiss says.
Specifically, images from a scanning electron microscope showed the lithia in the solid-state cell were deposited on the rough surface of the cathode in 500-nanometer-deep valleys. During the recharging of the battery, the lithia decomposed as the lithium ions flowed in the other direction and were redeposited on the lithium-metal anode.
The researchers believe that the battery’s ability to last through many cycles came from two sources: The reliable growth and decomposition of the lithia, and the solid electrolyte. Since there was no liquid in the battery to contact the cathode, there were fewer pathways for unintended chemical reactions compared to what goes on inside a liquid-electrolyte battery.
“That’s why we think that it can run quite longer: because you don’t get these side reactions, parasitic reactions, that destroy the electrolyte or the cathode surface,” he says.
The four-electron lithium reactions also account for the Asadi battery’s potential to store 1,000 Wh/kg of electricity. To put that achievement in perspective, Venkat Viswanathan, an associate professor of mechanical engineering at Carnegie Mellon University who is not affiliated with the research of Asadi and Curtiss, says that batteries capable of providing 1,000 Wh/kg would be “transformative” for short-distance aircraft. Halle Cheeseman, a program director at ARPA–E, the Advanced Research Projects Agency–Energy, sees 1,000 Wh/kg as a milepost for battery development — the point at which some shorter flights could become electrified.
The specific energy requirements for electric aircraft of the future will depend on how far they’re flying and how many people they’re carrying. While prototype advanced air mobility aircraft today are powered by lithium-ion batteries, the limited capacity of the batteries means they can power only short flights — tens of minutes long — carrying a handful of passengers. According to a paper co-authored by Viswanathan in Nature, batteries with 300 to 400 Wh/kg — at the upper limits of what lithium-ion batteries can provide — could power advanced air mobility aircraft for intracity travel. Commuter aircraft with up to 19 seats will need 1,200 to 1,800 Wh/kg, and 150- to 180-seat planes will need 1,800 to 2,500 Wh/kg.
But even with the solid electrolyte lithium-air breakthrough, Curtiss estimates it will take another 10 to 15 years of development and scaling up before lithium-air batteries can power aircraft. He bases his estimate on the development timeline for lithium-ion batteries, which were conceived in the 1970s.
“We’re going to try to get funding for scaling this up to the commercial level,” says Curtiss, while emphasizing that there is still more lab work to be done.
Ahead, the researchers will try to improve on the specific energy of the battery and increase the number of recharging cycles. In addition, they’ll try to lower the amount of electricity it takes to fully charge the battery, which is its charging efficiency, and speed up how fast it can discharge its electricity and then recharge, which is the charge rate.
As for scaling up, the issues for lithium-air are considerable but not insurmountable, says Cheeseman, who oversaw research and development for Rayovac Corp. in Madison, Wisconsin, in 2004, when the company tried to develop a lithium-
“The challenges that must be overcome aren’t necessarily more overwhelming than for any battery chemistry,” says Cheeseman, who uses they/them pronouns. “It just takes a long time to work through the challenges to get something that’s commercially viable.”
Researchers expect to encounter new problems as they scale up the technology. “It’s one thing to have it working as something the size of a postage stamp. It’s very different to have it working at the size of what would be needed in a plane,” they say.
Obtaining enough material is likely to be a serious issue. Consider a battery’s cathode, for instance. When making a lab-scale version, researchers have the luxury of making its material a gram at a time. Larger batteries would require kilograms of the material, and making lots of those will require freight cars full of it. Right now, “sometimes those materials are only available in spoonfuls,” Cheeseman says.
The history of today’s lithium-ion batteries suggests that the problem should be solvable. “Back in 1980 somebody said, ‘Here is a cathode material that might work,’ and it was available in grams. Today, it’s available in thousands, if not millions, of tons,” they say.
Viswanathan predicts that either lithium-air batteries or another battery technology he is developing, lithium/fluorinated carbon batteries, will power the longest-range electric aircraft in about a decade, optimistically. He figures on five years to work out the science, two or three years to scale up a battery-producing factory and supply lines — borrowing from Tesla’s recent quick scaleups of its factories — and then two or three years for FAA certification.
Three technologies will shrink the development timeline for today’s new generation of batteries relative to the lithium-ion generation, Viswanathan says. Machine learning now allows researchers to perform much more virtual testing on the computer before having to perform lab tests. Robots operating around the clock in the lab can mix and test more variations of battery materials, and at a faster pace, than is possible with human testers.
“In the design cycle, you want to test as much as possible on the computer because that’s cheap,” he says. “You want to do as much as possible in the lab before you have to go out in the field.”
And tools like X-ray tomography, in which the deflection of X-rays beamed at a chemical reaction are recorded over time, and electron microscopy, in which images are recorded by aiming electron beams at a reaction, permit researchers to take “movies” of the atoms to see precisely what’s occurring and what might be going wrong, Viswanathan says. “All of these things will accelerate the discovery and optimization process.”
Viswanathan’s team at Carnegie Mellon built a robot to test electrolyte candidates with no human intervention. Testing alternative chemistries — whether robotically or through computer simulations — is a lot like cooking. “You’re trying to find a recipe,” he says. “It’s a little like making a new sort of cocktail. You’re trying to get the ingredients right.”
As with cooking, a pinch or dash of an ingredient, or many ingredients, can make all the difference.
“Additives make or break batteries,” Viswanathan says. “They are all added in small proportions to give that one little thing.”