If True, QuantumScape Has Made the Biggest Leap in Batteries Since the Debut of Lithium-Ion
But most battery researchers don't believe it
Feb 2·11 min read
|
Onits first day of trading in November, shares of QuantumScape, a lithium-metal battery startup, surged by 57% in price. Then 10 days later, the price doubled, and less than two weeks after that, it was up another 72% — a total 5.7-fold increase in less than a month. The price has since plunged back to earth — sort of. As of the close of trading yesterday, it was up a mere 80% since its debut two months ago.
But the stock's dramatic rise has its logic if you understand that QuantumScape is at the center of a whirlwind in lithium-ion battery technology. For four-and-a-half decades, ever since the invention of the first, primitive lithium storage device at Exxon, researchers have sought but failed to achieve what the oil company couldn't: to make a battery powered by pure-lithium, the lightest metal on the periodic table. If they could, they would unlock immense power for electric vehicles — much more than the plain-Jane lithium-ion battery. But no one could navigate the stubborn metal.
The scourge was a sort of cancer — a nightmarish growth that tends to bloom from pure lithium metal while the battery is in use and trigger the device's death. The dreaded growth is called dendrites, and researchers have failed to rid batteries of them, and even to discover what precisely causes them. The one thing the research community has done is agree that dendrites look horrible — like the craggy root of a tree, as some say, or spooky tendrils, spiky branches, or crawling moss. Whatever is most apt, researchers know that when dendrites burrow in, they can bid the battery they are working on goodbye.
In December, though, QuantumScape came out of a decade of stealth with remarkable data for a lithium-metal-based battery: If scaled up, its cells would charge up every time in 15 minutes and still retain 90% of their original capacity after 240,000 miles of driving. In one test, its cells even charged up in two minutes. Delirious excitement followed QuantumScape's coming-out party on Wall Street and in the battery community.
But in an hour-long interview last week, two of QuantumScape's founders were most interested in talking about a single thing: How they had beaten dendrites.
CEO Jagdeep Singh and CTO Tim Holme said the company had discovered what precisely causes lithium dendrites, and that they had also developed a material that suppresses them before they grow. If validated, their breakthrough would be the biggest in batteries since the commercialization of lithium-ion itself in 1991. "We're lucky that in the end nature had a material," Singh said. "The team was able to discover it. And importantly, the team was able to manufacture it in a way that was low-cost and scalable. Because if you couldn't do that latter point, it wouldn't have mattered."
Over the years, QuantumScape has been extraordinarily secretive, refusing to divulge the barest details about what it has been doing. In September, it began something of a coming-out built around its debut on the New York Stock Exchange. Still, the company has steadfastly refused to say much of anything about its crucial separator material, the thin strip of film that lies between the two electrodes. (The consensus in the battery community is that it's a material called LLZO, a flexible ceramic that has attracted attention because it doesn't disintegrate when in contact with lithium metal.) This is the first interview that Singh and Holme have given to describe their experience making the separator.
One question is whether we are watching a "Bannister Moment": For 68 years, starting in the closing decades of the 19th century, the world's greatest coaches and runners made a concerted effort to achieve a milestone — the sub-four-minute mile. In 1954, a 25-year-old British medical student named Roger Bannister finally ran one. But his record was short-lived. Just 46 days later, an Australian runner beat Bannister's feat, and in the subsequent year, three men did so in a single race. In the years since, more than 1,400 more people have done it, too. It turned out that the barrier to the four-minute-mile was all in the mind.
So is the defeat of dendrites, too — if it is confirmed and scaled up — a matter of the mind? Will the rest of the field now beat them?
In2009, Singh got it in his mind to invent the better battery. He quit his job as CEO and founder of Infinera, a Silicon Valley optical networking company, and started incubating the idea. At Stanford, where he had gotten his master's in computer science, he met Holme and Fritz Prinz, a material science professor. The next year, QuantumScape was born.
"We didn't have any religion about how we would build the battery, only that we wanted greater density, faster charge times, greater safety, and so on," Singh said. To be cheaply manufacturable, the battery also could contain no rare-earth or scarce materials, like germanium, platinum, or palladium, and had to be continuously manufacturable, so off-the-shelf lithium-ion machines could be used.
The men quickly settled on lithium metal as the best chance of achieving this super-battery. But because the metal is so volatile, current EV batteries use just sprinkles of lithium that intercalate into a safe graphite anode. To make pure lithium metal useable, most researchers are attempting to develop a solid separator, and thus avoid the standard liquid electrolyte that makes lithium metal react so badly. Toward the same goal, Singh had raised around $150 million in venture capital, and he and Holme quickly assembled a company of around 100 engineers. Now, Holme and his team got to work drawing up a list of all the possible materials that might produce a successful separator.
"But every one of the systems we looked at was dendriting," Singh said. "That turned out to be the single biggest challenge." A year passed, two, then three. By 2014, Singh was despairing. So were some of the rest of the team. It was the dendrites. Singh shows me a slide. "This is a really scary picture," he said. "This is a solid separator but it didn't prevent dendrites. Not only can you see lithium metal that's made its way through the material. But these gray regions that you can barely see — they represent lithium that's actually inside the material. Lithium will literally snake its way in there, and even if it doesn't pop all the way out, it's this monster that's lurking. And with enough cycles, it will just burst out. It's like that monster in Alien in the '80s where there was an alien inside Sigourney Weaver and it pops out in the end."
Scary stuff. And nothing was working. "Those were some dark times, I gotta tell you," Singh added, "because you've raised the kind of capital you've raised and there's not a clear light in the tunnel. You don't know what the right course of action is. You have cash in the bank. Should you continue to invest it or should you say to investors, 'Guys, here's the money back. It's not going to work'?"
Holme pulled together all his engineers. "Time out," he recalls saying. "Put your work down. Everybody, stage left. We're going to start working on this dendrite problem. If we don't solve this, we don't have a product, we don't have a company." For a year, the entire engineering team worked only on dendrites. They collected every known theory about what causes them. First was that if the separator were hard enough, dendrites could physically not poke through and short circuit the battery. Only, that turned out not to be true, because lithium managed to knife their way through hard ceramics anyway. Then they looked at the theory that a soft separator would do the trick. It didn't work, either.
Finally, they discarded the literature. "We had to go back to first principles and come up with our own theory of what causes dendrites," Singh said. "And we had to actually develop our own metrology, which are measurement techniques to measure the quality of the material we were making because it turns out that some of the things that cause dendrites were not even measurable with normal metrology techniques."
In 2015, the team finally settled on an explanation for lithium dendrites. They began testing it on the materials they had gathered. Finally, they found one that policed the dendrites. It was great. "My personal depression started to lift," Singh said.
Only, they had spent five years to reach this stage, and all they really had was a tiny shard of material. They needed to make it larger and larger and better and better, with absolutely no defects such as pinholes that would attract dendrites. Each step up in size took six months to a year. In all, this phase required another five years. The last step, made in the end of 2019 and into last year, was taking the cell from 30-by-30 millimeters in size to 70-by-85.
So what did they discover? What causes dendrites? They aren't saying. "We may make it public eventually," Singh said. "But the industry doesn't know this yet and it took a lot of work and sweat and blood to get there, so we feel that if we share that it makes it easier for competitors to enter as well. So we're trying to avoid sharing that."
Given the profound trauma that dendrites have caused in the battery community over the decades, it's perhaps not surprising that not many researchers appear prepared to accept QuantumScape's claim at face value. When I called around to researchers, a typical question I got back was whether the company was talking about the discovery of a universal solution for lithium dendrites, or something more limited. "Can its solution be applied to other lithium-ion systems, or is it unique to its own system?" asked James Frith, head of energy storage at BloombergNEF, a renewable energy research firm. That is, even if it is validated, if no one else can benefit from QuantumScape's discovery, is it really a scientific breakthrough?
I asked Singh, the QuantumScape CEO, whether he was talking about a broad dendrite solution. "What our team discovered is applicable to solid-state separators, which to our knowledge, is the only way to prevent dendrites," he responded in an email. That seemed to suggest that QuantumScape's approach could work with the separators being produced by any of its solid-state competitors. If so, that could be the Bannister Moment. But, as with all of QuantumScape's impressive claims, confirmation is the operative word. And that "depends on getting independent people to verify it," Frith said. "That is the only way to truly confirm it." As of now, independent validation does not appear to be on Singh's priority list.
Jeff Sakamoto, a materials science professor at the University of Michigan whose battery work is similar to QuantumScape's, said that if what the company is reporting about dendrites is true, it is a substantial leap. "Lithium metal is the Holy Grail," he said. "That is the breakthrough." But Sakamoto is not convinced that the breakthrough is technological. Rather, from perusing the company's patents, he thinks that QuantumScape has simply devised a shrewd way to operate its batteries. Sakamoto specifically suspects that QuantumScape is "pulsing" its cells, or regulating the current going into the battery. "If they start to form dendrites, they can pull it back by reversing the current," he said. "If you carefully control it, you can get around the problem." Clearly, Sakamoto would be more impressed if he believed that QuantumScape had in fact made an advance in materials science.
"From a deep dive into their patent portfolio, there is nothing that stands out as an obvious breakthrough magic material or magic coating that enables fast charge," Sakamoto said. "There is no trail of evidence that would explain how they made a leap in performance, in other words. And, if they kept the breakthrough material or coating as a trade secret, it would be too risky since there would be a clear signature in their product. This makes me think it's more of how the cell is cycled than what's inside the cell."
The greatest reservation cited about QuantumScape's dendrite claims is that all of its assertions revolve around a single-layer cell, and not the eventual 100-layer cell that it has to create to go commercial. Most serious people in the industry say that scaling up is the hardest task in a successful battery —Elon Musk, the CEO of Tesla, for instance, has called scaling up "99.9%" of the work. I have spoken to no one in the battery community outside of QuantumScape's tight-knit group who is confident that the company can pull off the scale-up.
Unsurprisingly, that includes the company's competitors. Among them is Josh Buettner-Garrett, CTO of Solid Power, a solid-state battery-maker based in Colorado. He never mentioned QuantumScape by name, but throughout an interview made it clear he was talking about the company. Creating a single-layer cell is one thing, he said. "The big challenge comes in how do you achieve the same results in automotive scale large cells." Buettner-Garrett said, "They tend not to be able to point to how these [cells] will be produced with the same perfection at automotive scale at the required cost point."
Venkat Srinivasan, head of the Collaborative Center for Energy Storage Science at Argonne National Laboratory, suggested that the scaling problem could leave the way open for a rival electrode formulation, such as silicon, to commercialize first. "QuantumScape's innovation is scientifically huge, but I do wonder if it will be technologically huge," Srinivasan said. "In 15 years, we may end up saying silicon was instrumental in changing the EV landscape and moving us to full electrification."
With all this skepticism, I emailed Singh a last time. "You are saying that your team established the definitive reason why dendrites happen with lithium metal, meaning not another theory but the actual reason, correct? And second you are saying that your solution prevents dendrites — not that it impedes them under certain conditions but actually prevents them from forming?"
In an email back, Singh started out with what he might have thought was modesty, but if so, the effect was to dig in deeper. "Humility prevents us from saying we've found 'the definitive reason dendrites form,'" he wrote. "Just like it might have been hard for someone to say Newton's theory of gravity was 'the definitive theory' of gravity, even though it worked well enough experimentally allow the prediction of the paths of cannonballs and the timing of eclipses, etc. And indeed eventually Einstein came along and found what might be a better theory in general relativity. And of course, there might be an even better theory than general relativity waiting to be discovered that might better predict, for example, what happens in black holes, etc.
"What we'd be comfortable saying is we developed a theory of what causes dendrites, a theory that was different from the conventional wisdom, and we were able to empirically validate the theory well enough to allow us to build solid-state separators that work at record-high levels of current density without needing elevated temperatures while delivering >1,000 cycles," he said. "These are of course, the parameters you need to hit to be useful in real cars, so that's why this development is interesting and important."
When I have spoken with oilmen, I have gotten the sense of an animate actor — oil, so much of it sloshing around the world, making some countries rich, and most recently warming the Earth. For different reasons, Singh gave me a similar feeling about lithium. It, too, was alive. But now, it was possibly under a bit of control. "Working with lithium metal and with batteries has a way of producing humility," Singh said, "because you just don't know if lithium is going to find a way to get you."
Editor at Large, Medium, covering the turbulence all around us, electric vehicles, batteries, social trends. Writing The Mobilist. Ex-Axios, Quartz, WSJ, NYT.
Posts
No comments:
Post a Comment