Bottled Lightning

by Seth Fletcher

  True, Eberhard had been fairly belligerent in telling his side of the story. His rants on his Tesla Founder’s blog became famous for their bile. By the time this book was written, the blog, teslafounders.com, had been deleted, but in July 2008 Fortune quoted a “typical” Eberhard post: “The company has changed so tremendously since I started. It’s very secretive and cold now. It’s like they’re trying to root out and destroy any of its heart that might still be beating.”

  In any case, Tesla had blown through its launch date of August 2007. By the time the Plug-In conference began in San Jose, a quick drive down the freeway from Tesla headquarters in San Carlos, Tesla had delivered only seven “Founder’s Series” cars, to company intimates and big-time investors like Larry Page and Sergey Brin, along with Musk and Eberhard. Tesla’s strife drew its own spectators, and blogs like the Silicon Valley–oriented gossip site Valleywag.com began obsessively covering the company’s every misstep.

  Meanwhile, as GM’s fortunes declined, the Volt’s critics were relentless. A month after the Plug-In conference, Bill Reinert, manager of advanced technology at Toyota, told EV World that Toyota employees had a “death watch” going for the Tesla Model S, Fisker Karma, and Chevy Volt. The reason: the outrageous cost of the batteries. A public-relations rep attempted to walk back Reinert’s comments—“As a company, we do not have an official death watch anywhere,” she told the website Greentechmedia.com—but the sentiment was consistent with Toyota’s stance on the Volt since the car was revealed.

  At Plug-In, GM had dispatched Jon Lauckner to preach to the converted and fight back the hordes who believed the Volt was nothing more than a wall of smoke and mirrors designed to distract the world from GM’s financial catastrophe. Over breakfast the first day of the conference, he smiled mischievously when I asked him about the project’s many doubters. “The window’s closing on the skeptics,” he said. “And the only thing that’s going to be left at the end of the day is: Are we on time?”

  The next day Lauckner defended the Volt in a conference room the size of a football field, before a crowd of nearly seven hundred. He said he had planned on giving a general update on the Volt project and announcing a partnership GM was starting with a few dozen North American electrical utilities, but that changed, because there was a fresh piece of scathing criticism to respond to. Three weeks earlier, an op-ed in The Wall Street Journal had accused the Volt of being nothing more than a ploy to make the federal government feel a little fuzzier about bailing out GM. Headline: “What Is GM Thinking?” Lauckner had read the column while in China on business, and he was livid. He quickly e-mailed an angry rebuttal to several top executives and PR people marked “high” importance. “I won’t deny this raised my blood pressure,” he told the audience. “It’s absolute nonsense. It called into question whether the search for automotive technology that doesn’t involve petroleum is worthwhile.” The first PowerPoint slide to appear on the projection screen behind Lauckner showed the headline from The Wall Street Journal with a new byline: Jon J. Lauckner. The gag drew tepid laughter.

  “Pick your issue, and the common denominator is oil,” he said. “One fact stands out above all others: Going forward, we can no longer rely solely on oil to supply auto energy requirements.” Which alternative energy source is the answer, well, that’s an open question. “However, we increasingly believe the solution involves the electrification of the automobile as soon as possible. There has been a shift in the debate from ‘if’ to ‘when.’”

  Then, for some reason, he decided to probe at the old EV1 wound—old to him, perhaps, but still raw in this audience’s mind. “Some folks have recently suggested that we just dust off the tooling—if we can still find it—and crank up production of the EV1,” he said. “And look: the technology of the EV1 was state of the art ten years ago. But GM has chosen to put our efforts behind newer and better technology that will have greater functionality and therefore a much greater chance of high-volume marketplace acceptance.”

  That technology, obviously, was lithium ion, and when the Volt was unveiled a year and a half earlier, “critics voiced doubts,” Lauckner acknowledged. “Lithium ion was a dream. Even if it was achievable, we couldn’t bring it to market by 2010. Today I can tell you with a lot more development under our belts that we’re confident.”

  6

  THE LITHIUM WARS

  On a damp morning in November 2009, Yet-Ming Chiang was in prime form. In an expansive conference room in Boston’s Back Bay neighborhood, he was addressing the hive mind of the advanced battery research community—an audience of more than a hundred academic and industry researchers gathered for the fall meeting of the Materials Research Society. His half-hour talk was a tour de force of scientific pitchmanship. The fifty-one-year-old scientist was boyishly energetic, with the poise, confidence, and polish of a man absolutely sure of his work, a man who might be on the verge of making a seriously large amount of money.

  He sprinted through a series of slides that demonstrated how the exotic electrode powders that he and his underlings spent their time synthesizing in a lab across the Charles River scaled into the technology that could power a new generation of hybrid and electric vehicles. He explained how earlier that year, his company, A123 Systems, had installed the “highest powered lithium-ion battery available today” in a McLaren Mercedes Formula One “Kinetic Energy Recovery System”–class race car, and how in a subsequent race the extraordinary bursts of power that the battery delivered on demand helped the driver move from eighteenth out of twenty to finish in fourth place. He explained that today’s lithium-ion batteries are “highly mass and volume inefficient,” and that by mass more than half of the stuff in a battery was inactive, auxiliary material, electrochemical packing peanuts that Chiang would very much like to replace with more active electrode material. We can improve lithium-ion batteries by a factor of two just by using the space inside them more efficiently, he said. Thicker, denser electrodes, like the ones being developed in his lab, could help accomplish just that.

  He didn’t mention that A123 Systems is built on a disputed technology, the subject of interminable legal wrangling and scientific controversy. He didn’t mention the Canadian company that claims to have the exclusive right to manufacture and sell the kind of batteries that had made A123 one of the most famous among the new crop of American clean-tech start-ups. He didn’t have to, of course. Everyone in the room knew the story, or at least one side of it, quite well.

  Chiang also chaired the final session of the day, and afterward I hung around until the well-wishers and graduate students eager to introduce themselves to him had dispersed. We had been in touch by e-mail, and I walked to the front of the room to introduce myself. He asked me to remind him what, exactly, I was researching, and as soon as I got three words out he cut me off: “The lithium wars of the early twenty-first century!” he said with a grin. “I tell my students they’ll look back and be able to say, ‘I was there.’”

  A central battle in the lithium wars began in the early 1990s, a few years after John Goodenough left Oxford to take a post at the University of Texas at Austin. Sony’s commercialization of the lithium-ion battery had cemented John Goodenough’s reputation as the leader of his field. He may not have seen a penny in royalties, but in solid-state chemistry circles he had become well-known as the man whose compound had transformed portable electronics. In 1986 he had brought a postdoc over from Oxford named Arumugam Manthiram, and together they had established a solid-state chemistry lab in UT’s sprawling engineering school. In the late 1980s, during the high-temperature superconductor craze, their battery research faded in priority, but the Sony announcement retrained the scientific community’s attention on the subject. Soon scientists were searching for a cheaper and safer successor to lithium cobalt oxide, one that replaced an expensive, toxic material (cobalt) with a cheap, abundant, benign element like iron.

  In 1993, a visiting scientist named Shigeto Okada arrived in Goodenough’s lab. He was
a researcher at Nippon Telegraph and Telephone, otherwise known as NTT. Goodenough put him to work with two other scientists, a graduate student named Akshaya Padhi and a postdoc named Kirakodu Nanjundaswamy, studying a few variations on an iron-based compound Goodenough had worked on previously. “It was supposed to be a perfectly simple scientific study, not necessarily aimed at doing a battery at that point,” Goodenough said.

  At the end of the spring semester, Goodenough and his wife left for a sojourn at their house in New Hampshire, and while they were away Padhi spent his time studying a synthetic version of the mineral triphylite, or lithium iron phosphate. Surprisingly, it showed promise for use as a battery electrode. He had some luck putting it through “reversible” intercalation reactions—the atomic-scale burrowing that happens when lithium ions swim over to the electrode, dig inside, and dock there until the battery is recharged—and this meant that it was stable. The new compound also had the benefit of being made of nothing but cheap, almost free elements, which was exactly the kind of thing they needed to succeed the lithium-cobalt-oxide that Sony was now selling to the world. There was plenty of work left to do, however, and in the end, the results of their study were anemic. The new compound had a low capacity and was terrible at conducting electrons. Ions, no problem. Ions flew through this compound. Electrons were a different story, however, and a battery terminal in which electrons get bogged down as if in quicksand is useless. Nonetheless, the results were interesting enough that in 1996, Goodenough and Padhi decided to present their results at a meeting of the Electrochemical Society in Los Angeles.

  Michel Armand wasn’t planning to attend the conference, but when he saw the abstract for Goodenough’s paper, he knew he had to go to Austin. By then Armand was a visiting professor at the University of Montreal. He was also consulting for the energy company Hydro-Québec, which since 1978 had been doing R & D on a novel battery Armand invented in the early 1970s—a cell that used metallic lithium for the anode and a solid polymer to act as both the separator and the electrolyte. After leaving Bob Huggins’s lab at Stanford, he had gone back to France, and while there he had attempted to make iron phosphate into a lithium battery material, suspecting that it would make a good positive electrode for his polymer batteries. For technical reasons, he was never even able to synthesize the material, which is why when Goodenough announced that he and a student had made a batch and managed to get some interesting results, the fact that it appeared to be a relatively limp battery material didn’t matter to Armand. It existed, and worked at least a little bit. Armand was obsessed with the possibility of making battery electrodes based on cheap, almost infinitely available iron. “I immediately took the first flight out,” he said.

  Armand arrived in Austin with a small entourage to secure rights to lithium iron phosphate for Hydro-Québec. “He was very reluctant, because he didn’t believe in his compound,” Armand says of Goodenough. The trip was a success, however, and soon Hydro-Québec had an exclusive license on the technology, which meant that only Hydro-Québec—or a company that Hydro-Québec licensed the rights to—could legally manufacture and sell lithium iron phosphate electrode powder in North America.

  Within six months, Armand thought he had learned how to make the compound work. He believed that making particles of lithium iron phosphate that were each about the size of a particle of soot could solve the problem of low electronic conductivity. When individual particles “go nano,” or get down to the unfathomably tiny scale of less than a hundred nanometers wide, the particles are almost all surface area, and more surface area allows electrons to roam more freely. In the process of making those small particles, however, Armand’s group happened upon the second key to making lithium iron phosphate work. They started with a precursor material made of iron, phosphorus, and oxygen. Then they added a lithium compound and fired it. The burning of the lithium-containing compound ended up coating the tiny particles with carbon, and the conductivity shot up. “It solved everything,” Armand said. “The phosphate was perfect.”

  As it turned out, Goodenough, his student Padhi, and then Armand had developed something significant, a substance that would go on to be called one of the greatest materials-science advances of the decade. “But,” Armand says, “that was also the beginning of what would be—will remain—the biggest scandal in lithium batteries.”

  A few years after Michel Armand’s sprint to Austin, Yet-Ming Chiang’s group began working on “self-assembling” batteries, a far-horizon concept that means exactly what it sounds like. “We were trying to design into [different materials] the necessary attractive and repulsive forces to have a system in which cathode and anode particles assembled themselves,” Chiang said. To do so they were studying olivines, the class of compounds that includes lithium iron phosphate, which had properties that made them good candidates for that experiment. Chiang and his team were “doping” various olivines in an attempt to make them better conductors of electrons. Doping—adding tiny, targeted dashes of impurities to a material in order to tweak its electronic structure and therefore change its behavior—is one way materials scientists bend nature to their will. It’s how scientists can engineer the interior organization and behavior of electrons and atoms until, for example, what was a wafer of plain silicon becomes the basis for a microchip. When Chiang’s student Sung-Yoon Chung applied this technique to lithium iron phosphate, embedding niobium or zirconium atoms in just the right spots in the crystalline lattice, it seemed to cause an astonishing increase in the ability of the material to conduct electricity. It was like turning salt into metal. These were “very surprising results,” Chiang said.

  In October 2002, Chiang’s group published a paper that presented doped lithium iron phosphate as the next great hope for hybrid and electric vehicles. The paper argued that they had improved it on the atomic scale in such a way that it could make a battery cathode that could be completely discharged in three minutes, which is the kind of raw power that an electric-car battery needs. It was a breakthrough, Chiang’s paper argued, that “may allow development of lithium batteries with the highest power density yet.”

  Goodenough’s old collaborator Michael Thackeray, who was by then working at Argonne National Laboratory, wrote an accompanying editorial that emphasized the potential significance of Chiang’s experiment. This had “exciting implications” for “a new generation of lithium-ion batteries.” Thackeray acknowledged “one slightly controversial aspect” of the research: “that the olivine powders were synthesized from carbon containing precursors … Carbon can, of course, contribute significantly to electronic conductivity. Nevertheless, Chiang and colleagues carefully addressed this possibility and ruled it out.” His conclusion: “These results will spark much interest in the lithium battery community, who will undoubtedly want to repeat the experiments quickly to verify these very significant increases in electronic conductivity.”

  That was a bit of an understatement. The idea that adding a small number of metal atoms to lithium iron phosphate could transform it into a good electronic conductor generated considerable skepticism in the lithium-ion research community. To many, it just didn’t seem possible to transform this material so greatly with such a small tweak to its chemical composition.

  Michel Armand was incensed when he saw the paper. He believed there was no way Chiang’s method could have worked. To Armand, it was clear that Chiang had done essentially the same thing he had some years earlier—that in the process of preparing the material, he had unwittingly coated the particles with carbon. Instead of accepting that a carbon coat had made the material usable, however—which Armand had already demonstrated and presented publicly—Chiang simply clung to the more remarkable and useful explanation. Chiang’s paper was “false science,” Armand told me, something that should have never gotten past the independent scientists who judge studies before they can be published in a peer-reviewed journal like Nature Materials.

  “When the paper came out, I wrote to the editor and said, ‘I�
��m very sorry for you, because you’ve got a big scientific goof,’”Armand said. Armand began reproducing the experiments in Chiang’s paper, and soon he had formulated a rebuttal. Authored by Armand and two colleagues, the response was published as a letter to the editor in the 2003 issue of Nature Materials. The retort was delivered in the understated smack-talk of a scientific journal. “We suggest that the effects seen by Chung et al.”—Chiang’s student Sung-Yoon Chung was the first author on the paper—“are due to carbon for low-temperature samples, and to low-valency iron derivatives … It is beyond, not the scope, but the length of this letter to discuss the juggling of point defect chemistry equations to justify the results … . Unambiguously, it is the polyolefin worn from jars, subsequently charred into carbon, which is responsible for the good use of the LiFePO4 electrode.”

  To translate, Armand was accusing Chiang of either misunderstanding or misrepresenting his research. Chiang hadn’t doped anything, Armand argued. Instead, some of the lining from jars used in the experiment had been charred into carbon in the process of synthesizing the material, and that carbon had then coated the particles. There was a second fluke at work too, Armand argued. A metallic compound of iron and phosphorus (Fe2P) had also coated the particles, making it easier still for electrons to move around in the material. Together, these two lucky accidents made LiFePO4 into a fierce conductor of electrons.