Tiny metal deposits called dendrites threaten to curtail the development of rechargeable batteries. But engineers have solutions in sight.
Gasoline-powered automobiles seem destined for the rearview mirror. In March 2021, the Swedish company Volvo declared that by 2030 it will sell only fully electric cars. Just weeks earlier, Ford had announced plans to go all-electric in Europe by the same year, while GM is aiming for its cars to be fully electric by 2035. Last year, electric vehicles made up less than 3% of all new car sales in the United States, but a recent analysis by BloombergNEF predicts that their global market share will soar to nearly 60% in just 20 years (1).
Electric vehicles that can travel long distances and recharge quickly require safe batteries that pack a lot of energy into a small volume. Building those batteries, however, means overcoming a number of challenges. Chief among them is the problem of dendrites—disruptive, spiky growths of metal inside a battery that raise the risk of dangerous discharges.
Building a fast-charging battery that can suppress dendrites is a “holy grail problem,” says Yi Cui, an engineer at Stanford University in Palo Alto, CA. But recent findings suggest that the dendrite issue is not insurmountable. Experimental efforts to tame dendrites have produced promising, proof-of-concept demonstrations that capitalize on the strengths of lithium batteries while minimizing dendrite risk. These include strategies such as making nanoscale-level changes to the structure of the electrodes, studying the fundamental causes of dendrites, and exploring new materials for the anode-electrolyte interface and the electrolyte itself.
If these findings scale up, they will pave the way for a future in which fast-charging, long-distance electric cars are the norm rather than the exception. “We’re accustomed to pulling up to a gas station and fueling up to go on the next leg of our journey in a matter of minutes,” says Lynden Archer, an engineer at Cornell University in Ithaca, NY. “Unless and until we’re able to do something analogous with recharging, electric cars are going to be challenging for some customers.”
Over time, after much discharging and recharging, metal deposits pile up on a battery’s anode, its negative electrode. These deposits form structures that may look like porous moss or resemble needles. They may create narrow whiskers that reach away from the anode or larger growths that resemble tiny, metallic Christmas trees. Usually called dendrites—they resemble the neuron’s branching bodies of the same name—these tiny structures can cause big problems. Like weeds pushing up through a sidewalk, they can shoulder their way through the separator—a material barrier between the electrodes—and reach the cathode. This forges a connection that allows electrons to surge through the battery’s electrolyte, with potentially disastrous consequences.
In lithium-ion batteries, which power today’s electric vehicles, the electrolyte is often a flammable liquid. A 2017 report from the Federal Aviation Administration estimated that lithium-ion battery fires in computers, phones, or even e-cigarette chargers occur on flights about once every 10 days, and those fires can often be traced to separator problems that dendrites may have exacerbated. Dendrites have also been linked to other problems at the interface between the electrode and electrolyte that can deplete the battery’s capacity.
These dendrite complications tend to be worse in faster-charging batteries. “Avoiding dendrites is important so batteries can be charged at a high rate and ambient temperatures,” says materials scientist Nancy Dudney, who recently retired from Oak Ridge National Laboratory (ORNL), in Oak Ridge, TN, and has been at the forefront of energy storage and battery research for 40 years. She cautions that scientists don’t yet know all the ways that dendrites form. “There may be different mechanisms and multiple material properties that influence the tendency to form the lithium short,” she says.
Two Faces of Lithium
Since 1991, when the first lithium-ion battery hit the market, its design hasn’t changed much. Lithium is the near-ideal material for a lightweight, powerful, fast-recharging battery, because it’s the least dense elemental metal and each atom readily surrenders an electron. Most batteries in phones, laptops, and hybrids pack their lithium into a layered graphite anode. During charging, lithium ions exit the cathode and travel through the electrolyte to the anode, where they combine with electrons to form neutral atoms that sit between graphite’s layers. As the battery discharges, lithium ions leave the anode and return to the cathode, freeing electrons to generate a current.
The graphite anode helps to impede the formation of dendrites, but it also limits the amount of available lithium. Carmakers, chemists, and engineers warn that despite ongoing technology developments, the current lithium-ion battery will soon reach its maximum energy capacity and recharging speed. Even at their most efficient, today’s batteries will never power cars that can travel 1,000 miles on a charge, let alone an electric plane.
A better battery, they say, could have an anode of pure metal, with lithium as the most promising candidate. Liberating the metal from its graphite host could speed up recharging and offer an energy density of 500 watt-hours per kilogram or more, some two to three times the energy density of today’s highest-performing lithium-ion batteries (2, 3). “If you remove that dead weight of carbon, then you can get the highest possible energy density per unit weight,” says Zhenxing Feng, a chemical engineer at Oregon State University, in Corvallis. The implications are striking: An electric car that can travel 500 miles on today’s best lithium-ion batteries could travel, in theory, closer to 1,000 miles on the same number of batteries that use lithium metal.
The difficulty, though, lies in recharging those batteries. When lithium ions combine with electrons at the anode, they should ideally arrange themselves in neat, thin layers of atoms that optimize the limited volume available inside a sealed battery cell. But for a variety of reasons, lithium does not naturally settle into such an orderly configuration, especially during faster charging. When a battery charges too quickly, or overcharges, the metal piles up on the surface of the anode and begins to form dendrites. As the dendrites grow, Archer says, those porous, chunky structures provide more and more surface area for chemical transformations that inevitably lead to the proliferation of dendrites. There’s a razor-thin tipping point, too: The potential difference between storing lithium and plating it on the anode surface is just 200 microvolts. “We need to find a way to either get rid of dendrites during charging,” says Feng, “or completely suppress them.”
How to Block a Dendrite
In some ways, dendrite formation is similar to the practice of electroplating, in which a metallic solid crystallizes out of a liquid to cover a surface. One challenge, then, is to understand more precisely the process of electrodeposition that allows dendrites to form and branch in the first place. Researchers are also trying to identify metals that can be coaxed to crystallize in neat layers, which may make them suitable alternatives to lithium.
A team at the Pacific Northwest National Laboratory in Richland, WA, recently set out to better characterize dendrites by watching them form in real time. They used atomic force microscopy and an environmental transmission electron microscope to observe a tiny battery built just for the experiment and found that dendrites began to grow from the solid-electrolyte interphase, a thin film at the boundary between the lithium metal anode and the liquid electrolyte. The team’s video footage revealed a growing clump of lithium ions that quickly formed a thin whisker that protruded into the electrolyte. By modifying the recipe for the electrolyte, they were able to suppress this dendrite formation (4).
Archer, at Cornell, wants to understand how the physical design of the interphase, the material between a battery electrolyte and its anode, impacts the chemical reactions that drive dendrite formations. His group recently set out to study how variables such as the electrolyte chemistry, anode material, and the recharging current density affect the architecture of dendrites, taking a cue from the process of electrodeposition. They began by studying models of electrochemical cells that used zinc, instead of lithium, as its anode. Zinc is a common charge carrier in alkaline batteries, but there has also been recent progress in creating high-capacity zinc rechargeable batteries. When Archer’s collaborators ran a current through the electrolyte, it disassociated zinc ions that then accumulated on the anode and formed the telltale dendritic structures.
They ran similar tests on batteries using lithium and copper as the electrodes, varying the electrolytes and currents. What they found took Archer by surprise. Classical theory, based on electroplating experiments from the past few decades, predicts the rate at which dendrites form under various circumstances. But those results, he realized, were based on electrolytes with a very low salt concentration. Most batteries today use much more concentrated electrolytes.
When Archer’s group analyzed dendrite formation in salty, real-world electrolytes, they found that in these cases, classical theory failed to predict the inhomogeneous structures formed by dendrites. The findings suggest that dendrites form in a range of ways—both expected and unexpected—and in a variety of geometries and architectures. Notably, the researchers saw that the dendrites were preceded by highly porous, mossy metallic structures on the surface of the anode. “They were these kind of spongy, foamy structures that organized to produce a high surface area and then a low energy density,” Archer says.
Even more importantly, they found that constantly rotating the crystals in the anode solved the dendrite problem. Reorienting the crystals gave the aggregating metal atoms opportunities to better connect to the anode, and as a result the clumps didn’t give rise to whiskers, which prevented dendrite formation. “We could stop the process in its tracks,” Archer says. The results, published in June 2020, suggest that paying close attention to the crystal structure of the anode offers yet another promising approach to avoiding dendrites (5).
Work by a multi-institutional group including Feng, in Oregon, similarly suggests that tiny modifications to the anode architecture might hold the key to dendrite-free batteries. Feng’s team has been exploring how carefully designed, three-dimensional nanostructures on the surface of an anode could suppress the growth of dendrites during recharging. In January 2021, the researchers described a battery anode made of a porous zinc-magnesium alloy that could be charged and recharged over thousands of cycles (6). The patterned anode, Feng explains, suppresses dendrites by controlling reactions at the interface of the anode and the electrolyte.
“The key thing is the structure itself,” Feng says. “We found that by increasing the nanostructures on the anode, the [zinc deposits] can grow homogeneously.” His team also showed that the anode could function as part of an aqueous battery, particularly comptaible with seawater as the electrolyte. Seawater, he points out, is readily available and environmentally friendly, unlike the liquid electrolytes used today, which usually contain toxic, soluble salts.
In lab tests, their patterned anode suppressed dendrites over thousands of cycles and at a recharging current of 80 milliamps per square centimeter, which is many times greater than a lithium-ion battery can achieve in real-world settings. Next, his group plans to test the same approach on other metals and alloys, possibly including sodium and lithium. “What we’ve demonstrated is easy to assemble and seems easy to scale up,” he says. The main barriers are cost and figuring out how to capture the same benefits in marketable batteries.
“We knew that a lithium metal anode would be great, if possible. We also knew that either it would be our dinner, or we’d have to find new jobs.”
Modifying the electrolyte might also suppress dendrite growth. “You have to get a variety of materials and interface properties right,” says Dudney at ORNL. “The successful approaches will depend on the electrolyte being used, and the best choice is still debatable.” In October 2020, for example, Cui’s group at Stanford described how to implement a kind of “self-defense” system in a lithium-metal battery by adding a protein called fibroin to the electrolyte. Those proteins, they observed, attached not only to the surface of the lithium metal anode but also to the “buds” of newly formed dendrites (7). When a dendrite was completely covered by fibroin, incoming lithium ions didn’t attach to the end and instead formed neat layers rather than clumpy three-dimensional protuberances. By engineering the protein molecules to release slowly over the life of the battery, the group showed that the battery could withstand thousands of cycles of charging and recharging without significant dendrite growth.
A Solid Solution
But solid electrolytes could be the key advance—if technical barriers can be overcome. The idea is appealing: A solid, usually some combination of ceramic, polymer, or glass, would maintain a constant contact with the anode, which may help to suppress dendrites. A solid electrolyte would also be more durable than a liquid for real-world applications like electric cars. (It’s also not a completely novel approach: Michael Faraday experimented with solid-state batteries about 200 years ago.)
In 2019, Archer’s group reported on a liquid-solid hybrid electrolyte that combines a liquid’s advantage of easy ion transport with the dendrite-blocking potential of solids (8). And in 2020, Cui’s group described a solid polymer electrolyte that, unlike many other candidate materials, was fireproof (9).
Companies are already starting to take up the charge. In the fall of 2020, a startup called QuantumScape made headlines when it claimed that it had successfully built a solid-state lithium-metal battery that could power cars for long distances, recharge quickly, and not explode. The battery uses a solid electrolyte that, according to the company’s scientists, can tamp down dendrites. Data released in December 2020 suggested that the battery—which is still in the experimental phase—has a higher density than any other battery on the market and could recharge to 80% of its capacity in only 15 minutes (10).
The company has been chasing better lithium batteries for more than a decade, and engineer Tim Holme, the company’s Chief Technology Officer and cofounder, says that from the early days they were determined to find a new way to make pure lithium work. But they quickly hit the dendrite wall. “Early on, there was an especially intense period of about a year where we devoted unprecedented efforts toward solving this issue. We worked on it twenty-four seven,” says Holme. “We knew that a lithium metal anode would be great, if possible. We also knew that either it would be our dinner, or we’d have to find new jobs.”
After screening dozens of materials and running close to “half a million” simulations and experiments, Holme says, they found a way forward and developed a cell that, the company reported, can suppress dendrites in lab tests. Inside the cathode they added a gel catholyte—an electrolyte layer in the neighborhood of the cathode—and a flexible ceramic separator that acts as a solid electrolyte and suppresses dendrites in a metallic lithium battery. The battery is manufactured in an initially discharged state and without an anode in place, a design the company describes as “anode-free.” When it charges, lithium ions move across the electrolyte to form an anode in situ at the other end. Holme said that design saves space—the anode only takes up volume as an ultra-thin layer of pure lithium—and the company is now working on scaling up the production method. Their goal is to install high-performing solid-state lithium metal batteries in mass-market automobiles by the year 2024.
“An anode-free battery is worth exploring,” says Cui. But he remains cautious about early experimental findings, not only from QuantumScape but also from other companies developing and promoting solid-state batteries. “I think we’re still in the research phase and not ready for prime time.” Researchers have yet to improve the cycling efficiency of new batteries, he says, to avoid losing lithium to unwanted chemical reactions in the battery. Solid electrolytes can go a long way in suppressing dendrites, Cui says, but “it’s still not a completely solved problem.”
Archer, too, remains cautious. “Going anode-free doesn’t get rid of the dendrite problem,” he says, not least because lithium is so reactive it can accumulate in other locations to produce dendrites. In addition, ceramic polymers can give rise to unexpected reactions with other materials in the battery. Because QuantumScape has not shared details about the composition of its separator, some chemists remain cautious about interpreting the company’s results, Archer says.
There probably isn’t a single, ideal combination of electrolyte and anode structure in a lithium-metal battery that suits every application, Dudney notes. But perhaps multiple approaches will lead to a powerful, fast-recharging batteries that conquer the dendrite problem—and avoid other causes of shorts. Doing so could enable a future that includes more versatile electric cars, cheaper stationary storage for renewable energy sources—and even, Cui says, electric airplanes.