PALO ALTO, CALIFORNIA—On a drizzly, gray morning in April, Yi Cui weaves his scarlet red Tesla in and out of Silicon Valley traffic. Cui, a materials scientist at Stanford University here, is headed to visit Amprius, a battery company he founded 6 years ago. It’s no coincidence that he is driving a battery-powered car, and that he has leased rather than bought it. In a few years, he says, he plans to upgrade to a new model, with a crucial improvement: “Hopefully our batteries will be in it.”
Cui and Amprius are trying to take lithium--ion batteries—today’s best commercial technology—to the next level. They have plenty of company. Massive corporations such as Panasonic, Samsung, LG Chem, Apple, and Tesla are vying to make batteries smaller, lighter, and more powerful. But among these power players, Cui remains a pioneering force.
Unlike others who focus on tweaking the chemical composition of a battery’s electrodes or its charge-conducting electrolyte, Cui is marrying battery chemistry with nanotechnology. He is building intricately structured battery electrodes that can soak up and release charge-carrying ions in greater quantities, and faster, than standard electrodes can, without producing troublesome side reactions. “He’s taking the innovation of nanotechnology and using it to control chemistry,” says Wei Luo, a materials scientist and battery expert at the University of Maryland, College Park.
I wanted to change the world, and also get rich, but mainly change the world.
In a series of lab demonstrations, Cui has shown how his architectural approach to electrodes can domesticate a host of battery chemistries that have long tantalized researchers but remained problematic. Among them: lithium-ion batteries with electrodes of silicon instead of the standard graphite, batteries with an electrode made of bare lithium metal, and batteries relying on lithium-sulfur chemistry, which are potentially more powerful than any lithium-ion battery. The nanoscale architectures he is exploring include silicon nanowires that expand and contract as they absorb and shed lithium ions, and tiny egglike structures with carbon shells protecting lithium-rich silicon yolks.
Amprius already supplies phone batteries with silicon electrodes that store 10% more energy than the best conventional lithium-ion batteries on the market. Another prototype beats standard batteries by 40%, and even better ones are in the works. So far, the company does not make batteries for electric vehicles (EVs), but if the technologies Cui is exploring live up to their promise, the company could one day supply car batteries able to store up to 10 times more energy than today’s top performers. That could give modest-priced EVs the same range as gas-powered models—a revolutionary advance that could help nations power their vehicle fleets with electricity provided by solar and wind power, dramatically reducing carbon emissions.
Cui says that when he started in research, “I wanted to change the world, and also get rich, but mainly change the world.” His quest goes beyond batteries. His lab is exploring nanotech innovations that are spawning startup companies aiming to provide cheaper, more effective air and water purification systems. But so far Cui has made his clearest mark on batteries. Luo calls his approach “untraditional and surprising.” Jun Liu, a materials scientist at the Pacific Northwest National Laboratory in Richland, Washington, put it more directly: Cui’s nanotech contributions to battery technology are “tremendous.”
Making leaps in battery technology is surprisingly hard to do. Even as Silicon Valley’s primary innovation, the computer chip, has made exponential performance gains for decades, batteries have lagged. Today’s best lithium-ion cells hold about 700 watt-hours per liter. That’s about five times the energy density of nickel-cadmium batteries from the mid-1980s—not bad, but not breathtaking. In the past decade, the energy density of the best commercial batteries has doubled.
Battery users want more. The market for lithium-ion batteries alone is expected to top $30 billion a year by 2020, according to a pair of recent reports by market research firms Transparency Market Research and Taiyou Research. The rise in production of EVs by car companies that include Tesla, General Motors, and Nissan accounts for some of that surge.
But today’s EVs leave much to be desired. For a Tesla Model S, depending on the exact model, the 70- to 90-kilowatt-hour batteries alone weigh 600 kilograms and account for about $30,000 of the car’s price, which can exceed $100,000. Yet they can take the car only about 400 kilometers on a single charge, substantially less than the range of many conventional cars. Nissan’s Leaf is far cheaper, with a sticker price of about $29,000. But with a smaller battery pack, its range is only about one-third that of the Tesla.
Improving batteries could make a major impact. Doubling a battery’s energy density would enable car companies to keep the driving range the same while halving the size and cost of the battery—or keep the battery size constant and double the car’s range. “The age of electric vehicles is coming,” Cui says. But in order for EVs to take over, “we have to do better.”
He recognized the need early in his career. After finishing his undergraduate degree in his native China in 1998, Cui moved first to Harvard University and then to the University of California (UC), Berkeley, to complete a Ph.D. and postdoc in labs that were pioneering the synthesis of nanosized materials. Those were the early days of nanotechnology, when researchers were struggling to get a firm handle on how to create just the materials they wanted, and the world of applications was only beginning to take shape.
While at UC Berkeley, Cui spent time with colleagues next door at the Lawrence Berkeley National Laboratory (LBNL). At the time, LBNL’s director was Steven Chu, who pushed the lab to invent renewable energy technologies that had the potential to combat climate change, among them better batteries for storing clean energy. (Chu later went on to serve as President Barack Obama’s secretary of energy from 2009 to 2013.)
“At the beginning, I wasn’t thinking about energy. I had never worked on batteries,” Cui says. But Chu and others impressed on him that nanotechnology could give batteries an edge. As Chu says now, it offers “a new knob to turn, and an important one,” enabling researchers to control not only the chemical composition of materials on the smallest scales, but also the arrangement of atoms within them—and thus how chemical reactions involving them proceed.
After moving to Stanford, Cui quickly gravitated to the nexus between nanotechnology and the electrochemistry that makes batteries work—and accounts for their limitations. Take lithium-ion rechargeable batteries. In principle, these batteries are simple: They consist of two electrodes divided by a membrane “separator” and a liquid electrolyte that allows ions to glide back and forth between the electrodes. When a battery is charging, lithium ions are released from the positive electrode, or cathode, which consists of a lithium alloy, commonly lithium cobalt oxide or lithium iron phosphate. They are drawn toward the negatively charged electrode, called the anode, which is usually made of graphite. There they snuggle in between the graphite’s sheets of carbon atoms. Voltage from an external power source drives the whole ionic mass migration, storing power.
When a device—say, a power tool or a car—is turned on and demands energy, the battery discharges: Lithium atoms in the graphite give up electrons, which travel through the external circuit to the cathode. Meanwhile, the lithium ions slip out of the graphite and zip through the electrolyte and the separator to the cathode, where they meet up with electrons that have made the journey through the circuit (see diagram below).
Nano to the rescue
Cui and colleagues have applied several nanotechnology-inspired solutions to keep silicon anodes from breaking down and to prevent battery-killing side reactions.
Graphite is today’s go-to anode material because it is highly conductive and thus readily passes collected electrons to the metal wires in a circuit. But graphite is only so-so at gathering lithium ions during charging. It takes six carbon atoms in graphite to hold on to a single lithium ion. That weak grip limits how much lithium the electrode can hold and thus how much power the battery can store.
Silicon has the potential to do far better. Each silicon atom can bind to four lithium ions. In principle, that means a silicon-based anode can store 10 times as much energy as one made from graphite. Electrochemists have struggled in vain for decades to tap that enormous capacity.
It’s easy enough to make anodes from chunks of silicon; the problem is that the anodes don’t last. As the battery is charged and lithium ions rush in to bind to silicon atoms, the anode material swells as much as 300%. Then, when the lithium ions rush out during the battery’s discharge cycle, the anode rapidly shrinks again. After only a few cycles of such torture, silicon electrodes fracture and eventually split into tiny, isolated grains. The anode—and the battery—crumbles and dies.
Cui thought he could solve the problem. His experience at Harvard and UC Berkeley had taught him that nanomaterials often behave differently from materials in bulk. For starters, they have a much higher percentage of their atoms at their surface relative to the number in their interior. And because surface atoms have fewer atomic neighbors locking them in place, they can move more easily in response to stresses and strains. Other types of atomic movement explain why thin sheets of aluminum foil or paper can bend without breaking more easily than chunks of aluminum metal or wood can.
In 2008, Cui thought that fashioning a silicon anode from nanosized silicon wires might alleviate the stress and strain that pulverize bulk silicon anodes. The strategy worked. In a paper in Nature Nanotechnology, Cui and colleagues showed that when lithium ions moved into and out of the silicon nanowires, the nanowires suffered little damage. Even after 10 repeated cycles of charging and discharging, the anode retained 75% of its theoretical energy storage capacity.
Unfortunately, silicon nanowires are much more difficult and expensive to fashion than bulk silicon. Cui and colleagues started devising ways to make cheaper silicon anodes. First, they found a way to craft lithium-ion battery anodes from spherical silicon nanoparticles. Though potentially cheaper, these faced a second problem: The shrinking and swelling of the nanoparticles as the lithium atoms moved in and out opened cracks in the glue that bound the nanoparticles together. The liquid electrolyte seeped between the particles, driving a chemical reaction that coated them in a nonconductive layer, known as a solid-electrolyte interphase (SEI), which eventually grew thick enough to disrupt the anode’s charge-collecting abilities. “It’s like scar tissue,” says Yuzhang Li, a graduate student in Cui’s lab.
A few years later, Cui and his colleagues hit on another nanotech solution. They created egglike nanoparticles, surrounding each of their tiny silicon nanoparticles—the yolk—with a highly conductive carbon shell through which lithium ions could readily pass. The shell gave silicon atoms in the yolk ample room to swell and shrink, while protecting them from the electrolyte—and the reactions that form an SEI layer. In a 2012 paper in Nano Letters, Cui’s team reported that after 1000 cycles of charging and discharging, their yolk-shell anode retained 74% of its capacity.
They did even better 2 years later. They assembled bunches of their yolk-shell nanoparticles into micrometer-scale collections resembling miniature pomegranates. Bunching the silicon spheres boosted the anode’s lithium storage capacity and reduced unwanted side reactions with the electrolyte. In a February 2014 issue of Nature Nanotechnology, the group reported that batteries based on the new material retained 97% of their original capacity after 1000 charge and discharge cycles.
Earlier this year, Cui and colleagues reported a solution that outdoes even their complex pomegranate assemblies. They simply hammered large silicon particles down to the micrometer scale and then wrapped them in thin carbon sheets made from graphene. The hammered particles wound up larger than the silicon spheres in the pomegranates—so big that they fractured after a few charging cycles. But the graphene wrapping prevented the electrolyte compounds from reaching the silicon. It was also flexible enough to maintain contact with the fractured particles and thus carry their charges to the metal wires. What’s more, the team reported in Nature Energy, the larger silicon particles packed more mass—and thus more power—into a given volume, and they were far cheaper and easier to make than the pomegranates. “He has really taken this work in the right direction,” Jun Liu says.
Powered by such ideas, Amprius has raised more than $100 million to commercialize lithium-ion batteries with silicon anodes. The company is already manufacturing cellphone batteries in China and has sold more than 1 million of them, says Song Han, the company’s chief technology officer. The batteries, based on simple silicon nanoparticles that are cheap to make, are only 10% better than today’s lithium-ion cells. But at Amprius’s headquarters, Han showed off nanowire-silicon prototypes that are 40% better. And those, he says, still represent only the beginning of how good silicon anodes will eventually become.
Now, Cui is looking beyond silicon. One focus is to make anodes out of pure lithium metal, which has long been viewed as the ultimate anode material, as it has the potential to store even more energy than silicon and is much lighter.
But there have been major problems here, too. For starters, an SEI layer normally forms around the lithium metal electrode. That’s actually good news in this case: Lithium ions can penetrate the layer, so the SEI acts as a protective film around the lithium anode. But as the battery cycles, the metal swells and shrinks just as silicon particles do, and the pulsing can break the SEI layer. Lithium ions can then pile up in the crack, causing a metal spike, known as a dendrite, to sprout from the electrode. “Those dendrites can pierce the battery separator and short-circuit the battery and cause it to catch fire,” says Yayuan Liu, another graduate student in Cui’s group.
Conventional approaches haven’t solved the problem. But nanotechnology might. In one approach to preventing dendrite formation, Cui’s team stabilizes the SEI layer by coating the anode with a layer of interconnected nanocarbon spheres. In another, they’ve created a new type of yolk-shell particle, made of gold nanoparticles inside much larger carbon shells. When the nanocapsules are fashioned into an anode, the gold attracts lithium ions; the shells give the lithium room to shrink and swell without cracking the SEI layer, so dendrites don’t form.
Improving anodes is only half the battle in making better batteries. Cui’s team has taken a similar nanoinspired approach to improving cathode materials as well, in particular sulfur. Like silicon on the anode side, sulfur has long been seen as a tantalizing option for the cathode. Each sulfur atom can hold a pair of lithiums, making it possible in principle to boost energy storage severalfold over conventional cathodes. Perhaps equally important, sulfur is dirt cheap. But it, too, has problems. Sulfur is a relatively modest electrical conductor, and it reacts with common electrolytes to form chemicals that can kill the batteries after a few cycles of charging and discharging. Sulfur cathodes also tend to hoard charges instead of giving them up during discharge.
Seeking a nanosolution, Cui’s team encased sulfur particles inside highly conductive titanium dioxide shells, boosting battery capacity fivefold over conventional designs and preventing sulfur byproducts from poisoning the cell. The researchers have also made sulfur-based versions of their pomegranates, and they have trapped sulfur inside long, thin nanofibers. These and other innovations have not only boosted battery capacity, but also raised a measure known as the coulombic efficiency—how well the battery releases its charges—from 86% to 99%. “Now, we have high capacity on both sides of the electrode,” Cui says.
Down the road, Cui says, he intends to put both of his key innovations together. By coupling silicon anodes with sulfur cathodes, he hopes to make cheap, high-capacity batteries that could change the way the world powers its devices. “We think if we can make it work, it will make a big impact,” Cui says.
It just might help him change the world, and get rich on the side.