Bottled Lightning

by Seth Fletcher

  The insights gained from these studies allow scientists like Whittingham to puzzle out what, exactly, is happening inside these materials. Are ions traveling into the crystals and bonding where they’re expected to? Are they coming out when we coax the reaction into reverse? Is the structure breaking down as a result? This is the kind of work that, far downstream, determines comparatively proletarian things like zero-to-sixty times and highway passing power.

  And those are the things that get people like MIT’s Donald Sadoway excited about electrochemistry. “After you’ve driven an electric car, you don’t want to go back to internal combustion,” he said, “because all electric drive is just neck-snapping acceleration and flat torque curve—zero to thirty is neck snapping, fifty to seventy is neck snapping. It’s an exquisite ride. I’ve said for as long as I can remember that the problem with environmentalism in this country is that it’s been largely in the hands of the crunchies, and the mainstream American views it as sort of a penance, or a duty. People fail to realize that it’s a chance for a new start, a chance for reinvention and a way of making things—not just, ‘Well, we’ll compensate and we’ll try to make things as good as we have them now but with lower carbon intensity.’ It’s a chance to make things way better!”

  This line of conversation sent him into full pontification mode. “I teach this big freshman chemistry class here, and I do a little bit of editorializing from time to time, and I tell them to challenge themselves. Instead of thinking about a process to make steel that does minimal harm to the environment in comparison to the process we have right now, what about thinking about a process that actually cleans the air, and cleans the water, so that people fight to have the smelter sited in their neighborhoods because the trees are greener near the smelter, the water is cleaner having passed through the smelter? A car that drives down the road and its exhaust is purer than the air coming into the front of the car? Why aren’t you thinking this way? Why are you thinking that the best we can do is zero? Today we’re at a negative, and the best we can hope to achieve is to get as close to zero but on the negative side as possible? And I’m saying, why can’t you bust through the zero axis and go positive? Why not? It’s only the limits of your imagination that prevent it.”

  It’s impressive that longtime battery scientists, people who witnessed the stillbirth of electric-car revivals in the 1970s and the 1990s, can still muster that kind of imagination. Consider the case of Elton Cairns.

  Cairns led General Motors’s research into high-temperature electric-vehicle batteries in the late 1960s. He now holds posts at the University of California at Berkeley and Lawrence Berkeley Laboratory, where he specializes in the kind of next-generation lithium-battery technology that could go a long way toward meeting Lauckner’s goal of matching the performance of the gas-powered engine: lithium sulfur.

  Conceptually, the battery is simple. Composed of a lithium-metal anode, an elemental sulfur cathode, and an electrolyte made of a mixture of ionic liquids, liquid polymers, and a lithium salt, it works on the reaction of lithium and sulfur to form Li2S. Theoretically, the promise is immense: a lithium-sulfur battery should be able to store five times the energy of the lithium-ion batteries we use today.

  One of the immediate benefits is getting rid of carbon. The anodes in all of today’s lithium-ion batteries use LiC6, which by weight is

  only 10 percent lithium. Replacing all that carbon with active, charge-carrying lithium would be an instant advantage.

  Sulfur brings significant benefits as well. Because a sulfur atom is about a third of the weight of the cobalt oxide molecules found in today’s consumer-electronics batteries, the weight savings is almost as dramatic on the cathode side, particularly since one sulfur atom can bond to two lithium atoms. If a lithium-sulfur battery could be made to work correctly, it could store hundreds of watt-hours per kilogram, enough to jump up into the realm of the several-hundred-mile electric car. The big, fundamental problem is that sulfur is terrible at conducting electricity. “A lot of nanoscale engineering has to be done to make the sulfur electrode work well,” Cairns said.

  Sulfur-based lithium batteries may be years away, but there’s a near-term alternative that moves the technology in the same direction: silicon. Like sulfur, silicon can bond to more lithium atoms than can the materials currently used in lithium-ion electrodes. And lithium-ion batteries that use silicon are very close to production—like the cell that Panasonic will start selling in 2013, which will use a silicon alloy anode to reach an energy capacity of 4 amp-hours, a 30 percent improvement over the highest-energy cells available today.

  But silicon could make even greater leaps possible. In 2008, a young Stanford professor named Yi Cui attracted major attention when he unveiled silicon “nanowires” that could replace the carbon in the anode. “One silicon atom can bond with 4.4 lithium ions,” Cui said, whereas it takes 6 carbon atoms to bond with one lithium ion. Silicon is heavier than carbon, but even so, on balance, silicon can hold ten times more lithium than an equivalent amount of carbon.

  The catch is that when silicon reacts with lithium, it undergoes a huge change in volume, swelling when charged and deflating when discharged. Over time, this strains the electrode and causes it to break down. Cui believes his nanowires work around this problem by shrinking to sizes where mechanical strain is no longer a problem. “If the object is already smaller than the smallest thing you can break,” Cui said, “they don’t break anymore.”

  Cui, who in 2008 cofounded the company Amprius to commercialize silicon nanowires, is a sort of battery savant. In addition to his silicon research, he’s behind some of the oddest and most creative battery research happening—schemes to make lithium-ion batteries out of things like paper and fabric. Cui has shown, for example, that paper can act as a substrate for a lithium-ion battery. Basically, you take a piece of printer paper, dip it in an “ink” of carbon nanotubes or nanowires, and the paper becomes astonishingly conductive. “Paper is very light, and its internal structure is made of cellulose fibers, so it soaks [the active battery material] up like ink. Once you put the materials in, they can be accessed by the electrolyte very fast for high power.” He envisions one day using a factory akin to a paper plant to manufacture batteries.

  Donald Sadoway has a different vision, one that gives a glimpse of how the battery could evolve without lithium. A few years ago, he started thinking about the challenge of grid storage and the main limitations of batteries in that application. “Batteries intrinsically store charge, but they don’t like high current,” he said. “So the gambit for everybody else is, How do we engineer a battery that can handle high current?” Hooked to the grid, they’ll need it—to quickly soak up a load of electricity from a fast-running windmill, to quickly dump it into the system when the wind dies down and everyone in the neighborhood turns up the air conditioner.

  “What I did was turn the problem around. Why don’t I start with a device that intrinsically handles high current and then teach it how to store charge? I know what likes high current: an aluminum smelter. Now, how can I turn this into a battery? What is it about an aluminum cell that fails to make it a two-way street?”

  A student of Sadoway’s, David Bradwell, did a study of the concept for his master’s thesis. “We ultimately came to the realization that we needed liquid metals at both electrodes,” Sadoway said. “And that’s impossible. Because you make metal at the cathode and you make nonmetal at the anode. That’s where the invention came into play.”

  He grabbed a laminated, place-mat-size periodic table and laid it on the coffee table in front of us. “Over here you have the metals,” he said, pointing in the general vicinity of iron, “and 75 percent of the periodic table is metals. And over here are the nonmetals—the fluorine, the chlorine, the oxygen, the nitrogen. But it’s not a sharp departure from metal to the nonmetal zone. Along this staircase we have what we call the semimetals, or the metalloids.”

  He shifted from chemistry-lesson mode to recounting
his lightbulb moment. “So I was sitting looking at the periodic table one day and I was thinking, you know, antimony’s over here, and it’s got an electro-negativity of 2.05, which is sort of in the same vicinity as sulfur. And sulfur’s a nonmetal. But antimony’s a good metal. If you see a block of metal, it’s shiny and it conducts electrons. But if you take magnesium, which is way over here”—he points—“it’s got electro-negativity of 1.3. It’s such a good electron donor that if you put magnesium in the presence of antimony, antimony will be intimidated into becoming an electron acceptor. And that was when the lights went on.” He had found a loophole.

  He realized that antimony and magnesium have similar melting points but vast differences in density, which means that if you melt them both and put them in a bucket, they’ll separate. “So now I’ve got liquid metal, which is insoluble in a liquid salt, which is insoluble in a liquid metal. I’ve got the three layers. They phase separate. They stratify by density. And I don’t need any solid separator! It’s salad oil and water.”

  Aside from the general intriguing weirdness of building a battery out of a layer cake of self-separating molten metals, the great thing about this design, Sadoway said, is that it should, in theory, be eminently scalable. “If you want to scale a sodium-sulfur battery to the size of a thirty-three-gallon garbage can, you can’t,” because in order for a sodium-sulfur battery to work, the solid beta-alumina electrolyte—the same thing that triggered the rise of modern battery science in 1967—has to be paper thin. Small sheets of beta-alumina can be paper thin, but large ones can’t, because they’ll break and cause the whole system to fail. This liquid metal design, however, didn’t require any delicate paper-thin parts. These liquids should simply fall into place, no matter how many gallons of them are involved. “If I want to make something the size of a thirty-three-gallon trash barrel, I just make something the size of a thirty-three-gallon trash barrel. I’ve got liquid metal on the bottom, molten salt, and liquid metal on the top. And current leads in from the top and out the bottom, and away you go.” That, at least, is the vision. “Right now we’re at the size of a little coffee cup.”

  In 2009, Sadoway received a $7 million grant from ARPA-E, the agency in the Department of Energy that funds high-risk, high-reward research, to determine whether his vision for a battery can work at a large scale. He said that he doesn’t care about getting money or fame or a career boost out of the project. In his view, grid storage is an environmental necessity, but the market that any grid storage device will have to compete in is brutal.

  “Your competitor is not another battery” when it comes to grid storage, he said. “Your competitor is a gas-fired peaking unit. When the sun isn’t shining and the wind isn’t blowing, guess where your electrons are coming from? A gas-fired peaking unit. It’s a combustor. And I’m saying that instead, if we had a battery, we would have this beautiful situation in which you could draw electrons from the sun even when the sun isn’t shining. And that’s really compelling. That gets people out of bed in the morning.”

  What gets quite a few scientists out of bed each morning is the highest-risk, highest-reward battery technology of all: lithium-air. At the moment, lithium-air appears to be the best chance battery scientists have to beat gasoline. It is elegant in concept and, theoretically at least, extravagantly energetic.

  For the sake of comparison, consider that a lead-acid battery can store something like 40 watt-hours of energy per kilogram. Today’s best lithium-ion batteries can hold about 200 watt-hours per kilogram, and lithium-ion has a theoretical maximum of 400 watt-hours per kilogram. Lithium-air has a theoretical maximum of 11,000 watt-hours per kilogram. Even after handicapping to take into account weight, efficiency, and other foreseeable technological obstacles—after assuming that, for the sake of argument, the lithium-air battery will be able to deliver only 15 percent of its theoretical energy capacity—it still matches what gasoline, because of the terrible efficiency of the internal combustion engine, can deliver. And that is why scientists have been dreaming about it for decades. “As with all things in life where there’s a big prize, it’s not an easy one to reach,” Peter Bruce said.

  Lithium-air is probably the purest and earthiest battery chemistry possible, because in its simplest formulation it involves nothing but lithium, oxygen, and carbon—the lightest metal in the universe and two essential elements of all living beings. “You take the positive electrode of a lithium-ion battery and you replace it with porous carbon,” explained Bruce, who today is one of the world’s leading lithium-air researchers. “The electrolyte”—this could either be an organic solvent as in today’s lithium-ion batteries, a combination of polymers, or maybe even something based on water—“floods the pores of the carbon. Oxygen comes in from the air.” And then the lithium ions, the oxygen, and the electrons routing around through the external circuit all combine to form lithium peroxide (Li2O2), a solid. Then, as with any rechargeable battery, the

  whole thing happens in reverse. “When you charge up the battery, you actually decompose this solid material. It goes back to lithium ions and electrons and pumps oxygen into the atmosphere again.”

  Mainly because of the signal it sends to the world—IBM is interested!—the highest-profile lithium-air project right now is Battery 500, a lab dedicated strictly to lithium-air research at IBM’s Almaden Research Center. “A practical electric car will need a lot more mileage than is possible with lithium-ion batteries,” said Winfried Wilcke, head of the project. “Five hundred miles is the target you really want.” That, along with a nice resonance with the Daytona 500, is why IBM decided to call its lithium-air project Battery 500—“to differentiate this from incremental improvements of lithium ion.”

  The IBM project is taking a supercomputer-driven, fundamental-physics approach to the problem. “Electrochemistry has had a long history of a very Edisonian approach,” Wilcke said. “But for something as risky and difficult as a lithium-air project, that’s not good enough.” It’s risky and difficult because “wherever you look there are challenges,” he said. “It’s like climbing Mount Everest.”

  First there’s the maddening difficulty of recharging. Getting the discharge reaction to happen once, to get the lithium to react with oxygen to form solid lithium peroxide—thanks to recent advances, that part is not a problem. What is a problem is getting that reaction to happen in reverse, to get the solid lithium peroxide to decompose into oxygen and then plate pure lithium back on the negative electrode with mirrorlike smoothness, rather than coating it with the fuzzy metallic spikes that have long been the curse of lithium-metal electrodes. Power is another problem. The reaction between oxygen and lithium is intrinsically slow, far too sluggish to blast a car up the highway in a passing maneuver.

  Hope for the power problem comes in the form of nanotechnology, which, as it does for many other lithium-based battery technologies, increases the surface area of the individual electrode particles, thereby dramatically increasing the rate at which the battery can charge and discharge. (Catalysts can also help that reaction happen more quickly.)

  As with anything involving metallic lithium, safety is a concern, although Wilcke argues that we should first find out whether lithium-air batteries are even remotely feasible before worrying about safety issues. “A lot of people like to talk about dendrites and anode and danger and so on,” he said. “I kind of think this is a secondary project. For one, there is no need really to use metallic lithium in a lithium-air battery. One could use a carbon intercalation anode or something like Yi Cui’s silicon battery. You could combine that with an air cathode, maybe get up to 1,000 watt-hours per kilogram, but not to 1,700.”

  Jeff Dahn, the battery field’s most experienced spokesperson on the dangers of lithium metal, is still cautious. “What Moli Energy found back in the late 1980s was that lithium-metal electrodes, just under normal use, led to cell failures that were at just too high an incidence rate to make it a viable business,” he said. Nonetheless, Dahn urged me to ta
lk to Steve Visco, who as chief technology officer of the Berkeley company PolyPlus is in charge of making lithium metal safe and usable.

  PolyPlus was spun out of Lawrence Berkeley National Laboratory in 1990 as a sort of think tank for lithium-sulfur research. “In many ways it operated as a kind of innovation center for batteries,” Visco said. He told me that the company did “all of the groundbreaking work in lithium-sulfur chemistry.” As they were studying lithium-sulfur batteries, they found that they couldn’t find a way to stop the sulfur from interacting, undesirably, with the lithium. “There was only one real way we could see to stop that, and that would be to somehow encapsulate the lithium with a conductive solid electrolyte, like a thin glass layer.”

  After doing some basic research, they started looking for an existing material they could use for that protective layer. They were lucky. A company in Japan called Ohara was making exactly what they needed. “I called them and had them ship us some plates, and when I talked to their representative, he said, ‘Well, I have some of these plates, but they’ve been sitting on my desk for a couple years.’” One of the major challenges with fabricating a material like this is making it stable enough to sit on a desktop without reacting with the moisture in the air and corroding. “And I thought, ‘Wow, if they’re that stable, that they can sit on his desk for two years without turning into a puddle, I want to look at those.’ So I said, ‘Send me the samples that have been sitting on your desk for two years.’ He did, and we immediately put lithium up against those plates after actually verifying it was conductive, and it degraded. So we said, ‘Okay, that’s why nobody’s using it—it’s not stable against lithium.’”