The Powerhouse: Inside the Invention of a Battery to Save the World

by Steve LeVine

  A fight broke out between Argonne’s lawyers and its industry partners on one side and battery inventors on the other. Chamberlain was in between. He was prepared to grant the companies first rights to any inventions. But the reward to the inventors had to be specified, too. In a meeting with company executives, Chamberlain said the scientists were accustomed to public sponsorship, embedded in which was the incentive of royalties. It was like the lottery—they had no sense of entitlement to royalties per se, but they would not part with at least the potential of an outsized payday. Yet that unassailable aspect to lab work would vanish with a change from public to private sponsorship. That was because to industry, the calculus of success was measured by company triumph. In the case of a venture such as the Hub, Dow, Johnson Controls, and the rest would be seeking a particular profit formula that involved certainty. It would allow them to calculate how much they would need to charge for a product five to ten years before its actual manufacture and thus assess in advance the wisdom of a given investment. That this was a rare moment when they were petitioning Argonne for its help did not curb their demands. If Chamberlain could provide that certainty, the companies could live with a formula that satisfied the inventors as well.

  “The inventors have an implicit conflict of interest,” a lawyer told Chamberlain. “They should not be involved in any way or shape or form in any kind of decision.”

  The assertion that only the scientists, and not the companies, were self-interested was odd. It reflected a contradiction. The scientists were paid to research for the good of the nation and could gain personally in the remote case that an invention struck gold. But the relationship was symbiotic—American companies were increasingly dependent on the national lab system. They had a stake in keeping the inventors happy.

  Over the years, blue-chip companies had become proud glorifiers of tinkering. Silicon Valley no doubt thought it was living up to its reputation for inventiveness, but a more beautiful and useful smart phone or tablet wasn’t equivalent to the invention of the light bulb. Excluding its universities, the Valley was not producing Nobel Prize–level work, something that Bell Labs accomplished seven times.

  The Valley had yet to concede the hollow place it had become, but industrial companies had—they knew they lacked a crucial dimension: they could no longer hope to make the big leap on their own terms. That recognition underlay Dow Chemical’s eagerness to join En-Caesar or the Hub. It hoped that, in collaboration with Argonne, it would invent and commercialize the next revolutionary product.

  Argonne just as eagerly embraced this company interest, which could be a source of scarce research dollars, independence, peer respect, and validation. But did that mean, as the lawyers argued, that the benefits ought to be tipped in the companies’ favor? That they should play no role in determining their own compensation?

  “They have no say in it at all,” the Argonne lawyer replied.

  “Except they can say if they want to work or not,” Chamberlain said.

  “And you say, ‘Do you want a job or not?’” the lawyer said.

  “Can I? I can’t force them to do work.”

  “If they don’t do the work, then you put that on their evaluation and use the evaluation to push them out,” the lawyer said.

  Chamberlain stared impassively. “That is terribly unrealistic for a star scientist who already has a hundred and fifty percent more work than he can handle,” he said. “We just have to figure out a way to properly incent employees without creating some kind of nasty competition.”

  “They won’t know what the terms of the contract are.”

  “That’s true,” Chamberlain said. “They’re not supposed to.”

  Ultimately they would settle on a formula providing a third of the money to the institution where the invention was created; each Hub member would decide for itself how much to cut to the inventing scientist.

  Chamberlain had been alone in protecting the scientists’ backs. He bent to the assertion that the scientists were subordinate in the pecking order but secured their right to fight within their labs—where each probably exerted the most influence—for just IP compensation.

  He focused on the big picture, which he said was national and geopolitical. He wanted the United States to prevail in the battery race. To do so, the United States had to develop a manufacturing supply chain, a necessity if you were thinking of dominating an industry. That meant licensing IP and supporting capable start-ups like Envia, in addition to creating a new invention system to help remake the commercial capabilities of incumbent giants like GM and Dow Chemical.

  26

  “There Is a Problem with Your Material”

  Until Sujeet Kumar paid a visit to Argonne in 2007, the NMC in a way did not exist. It was only when he began pressing for a license, congratulating Thackeray for creating what he called the only cathode with a chance to challenge oil, that the South African—that anyone at the lab—grasped how exceptional the invention was.

  It was the nature of basic science labs. When Goodenough and Thackeray made their early back-to-back achievements—the former’s lithium-cobalt-oxide in 1980 and the latter’s spinel a year later—there was no market for their advances. Rechargeable lithium-ion batteries became commercial products only a decade later. When Sony commercialized Goodenough’s battery in 1991, it became the go-to formulation for virtually every laptop, smart phone, recorder, or really any battery-enabled consumer device. Goodenough’s batteries lasted longer than the technology they superseded—nickel metal hydride—and did not suffer nearly the severity of capacity loss after long use. Even two decades later, lithium-cobalt-oxide batteries remained the world’s workhorse consumer battery.

  But all of that was later. In contemporaneous time, the drive behind both Goodenough’s and Thackeray’s breakthroughs was little more than “academic curiosity.” No company showed the slightest interest in their inventions. Not that they never contemplated commercial application—they did; in strolls and hallway conversations on the Oxford campus, they shared their hopes for market adaptation. But the personal payoff became a citation in others’ papers and continued funding.

  The inspiration to use lithium-ion to revive electric cars, though, came later still. Lithium-cobalt-oxide was too expensive—specifically the ingredient cobalt—for serious contemplation in passenger vehicles. It packed a wallop of energy density—the best among any commercial battery—but was economically feasible only for compact purposes, meaning small electronic devices. When Toyota pioneered the modern-day push into electrics in Japan in 1997, its Prius hybrid again contained nickel-metal-hydride batteries.

  No other lithium-ion formulation was ready, either, well into the 2000s. Thackeray’s spinel bested lithium-cobalt-oxide in the laboratory, but in actual use it quickly degraded; it would be the mid-2000s before the degradation problem was resolved and spinel could be used in the 2011 Volt.

  It was the same with Thackeray’s NMC. He and Chris Johnson were well on their way to their breakthrough invention by the time of the first Prius, but they were in no rush because they knew of no interest in lithium-ion for automobiles. Again, the motivation was personal inquisitiveness.

  After their flush of triumph on patenting the NMC first, they were followed by Dalhousie’s Jeff Dahn and the New Zealanders. Colleagues congratulated Thackeray and Johnson; peer recognition grew. Even then Thackeray did not sense a world-beating breakthrough; he told Argonne’s intellectual property team to seek an American patent, but not to bother with the expense of an extraordinary protective effort. Do not file for international patents, he advised Argonne’s licensing team. Having missed out on the original U.S. patent, 3M, possessing the rights to Dahn’s version of the NMC, filed for international rights.

  But NMC 2.0—the tweak on the original composition—excited a stir in the battery community. It was identified as the next great battery material, superior to another futurist rag
e, lithium-iron-phosphate, a commercial version of which was created by A123, the celebrity battery start-up. A123’s chemistry—built on Goodenough’s original breakthrough—dispensed a burst of energy that was perfect for power tools. But because it worked at an average of just 3.3 volts, it underperformed when it came to endurance—the ability to stay charged for long distances, a quality sought by carmakers. Until now, automakers appeared prepared to adopt it anyway. But NMC 2.0, operating at higher voltage, provided improved endurance, threatening A123’s perch.

  A123 was not easily ceding the top spot. Bart Riley, its muscle-bound chief technology officer, had obtained samples of NMC 2.0. If it truly was revolutionary, Riley was prepared to make peace and figure out how to adapt to the new market. But he told his research team to run it through all the tests.

  From the company’s Waltham, Massachusetts, laboratory, Riley received an e-mail from a forty-one-year-old South Korean staff researcher named Young-Il Jang. NMC 2.0, Young said, appeared to have a problem. And not just any problem, but one so substantial as to possibly doom it outright for use in cars.

  Young told Riley and other colleagues copied in the e-mail that the jolt of voltage that gave NMC 2.0 its potency also seemed to thermodynamically change it. When the high voltage forced much of the lithium to begin shuttling, thus removing the cathode’s pillars, the structure sought to shore itself up and keep its shape. Other atoms rearranged themselves. Nickel took the place of lithium, and cobalt of oxygen. When the lithium returned, its old places were occupied. It had to try to find a new home. Thermodynamics made the atoms seek a new natural balance. The voltage steadily declined.

  Hence in actual application in an automobile, NMC 2.0 might not provide the consistent potency suggested when Thackeray was working on coin-size test cells in the laboratory. Unless the atomic reorganization could be controlled, Young concluded, the material might never find use in a car, which required reliability. In a gasoline-driven vehicle, the driver expected the engine to deliver more or less the same propulsion each time the accelerator was depressed—the pistons had to push out a smooth flow of power continuously, every time. It could not deliver the acceleration of a Ferrari the first day and a Mini Cooper on the hundredth. Similarly, in an electric system, the voltage in the second cycle could not differ from that of the fiftieth; you could not create a dependable, ten-year propulsion system with such instability.

  While perhaps bad news for Thackeray and Sujeet Kumar, Young’s find enthused Riley. He had heard plenty of passionate discussion of NMC 2.0, but not a word about this potential problem, what came to be called “voltage fade.” In Riley’s view, that created a substantial opportunity for A123. Over the years, no private battery company had seemed as bold, shrewd, and quick as A123. In this case, if Young was in fact the first researcher to notice voltage fade, he—and A123—could possibly be the first to solve it. A123 could patent the answer and then reap the profits of enabling NMC 2.0 for the electric-car world.

  Riley put in a stream of calls, chiefly to Jeff Chamberlain and Sujeet Kumar. To each he said the same: What were they doing at the moment to solve voltage fade? “Let me check,” Chamberlain said. Kumar answered similarly. Clearly neither was on top of the problem.

  At a battery conference in Montreal, Riley ran into Thackeray. “You know there’s a problem with your material,” Riley said. “It could be a doom factor.”

  “It’s something we have to address,” Thackeray said. In fact, the voltage question had recently been raised in the Battery Department, but until running into Riley, Thackeray did not consider it important.

  Back at Argonne, Thackeray checked with Sun-Ho Kang. A few months earlier, Kang, the South Korean who later would go to Samsung, had noticed unsteady voltage in NMC 2.0. Like his counterpart at A123, he thought that if voltage fade could not be stopped, you could not use the NMC 2.0 in a car. The Volt did not suffer from the problem. It used the original NMC, which did not undergo the same supercharging; a maximum of 4.3 volts was applied to it, lower than the 4.5 volts that activated NMC 2.0’s atomic-level chaos.

  Riley was suggesting that the parade of companies that had paid to license NMC 2.0—not just Envia, but BASF, GM, LG, and Toda—were holding a seriously flawed product. As his researcher had stated, NMC 2.0 perhaps could not be deployed for the purpose for which it had been purchased—longer-range, cheaper electrified vehicles. At least in its current state, it perhaps could only be used at lesser voltages, which would mean performance not much different from the lithium-cobalt-oxide batteries commercialized two decades before. There might be no reason for anyone to absorb the expense of switching to NMC 2.0.

  • • •

  If you asked the battery guys at what stage they understood that there was a problem with NMC 2.0, it prompted a nervous response. They would go quiet, glance around, and provide not quite precise answers. This conveyed the impression that either no one knew the precise answer or no one wanted to disclose it. The reason being that, if you looked at the situation squarely, you could not escape the conclusion that Argonne had in fact sold the companies a faulty invention. Not that the companies themselves were off the hook—the engineers, venture capitalists, and other executives and staff who had signed off on the licenses had to be in some hot water among their bosses, too. If anyone was predominantly responsible, it was the Thackeray team, because their names were on the patent.

  Chamberlain, who had led the negotiations on Argonne’s behalf, said simply, “We didn’t know about it.” But how was that possible? “Because making a product is not the scientists’ objective. You have to look at a certain data set to notice the fade,” he said. “If you look at a different data set where all of your requirements are for capacity, you can actually miss the voltage curves.” He added, “That is why interaction with industry is so important, because if you are making a product, like a battery that is going into a car, you look at everything like this.”

  What Chamberlain described was precisely why Riley’s team might manage to steal a march on the industry. The reason why no one had picked up on voltage fade was that it had been a problem in no other major battery chemistry. So when battery guys evaluated this new chemistry, they skipped the voltage. “There was a blind spot,” Riley said.

  Papers Thackeray and Kang had published as far back as 2007 revealed weaknesses of NMC 2.0. The papers included charts revealing the fade phenomenon. Thackeray argued that that meant that Riley had stumbled over nothing new. But he was wrong. Thackeray and Kang published charts of voltage fade without explaining its significance, because they themselves did not grasp that it was a potential showstopper for the NMC 2.0. Riley did.

  As for why Young-Il Jang did check for it, Kang said he himself tipped off his A123 counterpart. But if so, Young-Il was already prepared to find the fade because he had observed a similar phenomenon in a cathode on which he worked for his doctoral thesis in the late 1990s. That allowed Young and Riley to move faster to report the finding.

  Whatever the case, Riley, armed with what he called a major scientific “scoop,” now pursued talks with both Chamberlain and Kumar. With the latter, Riley held out a seductive offer—A123 and Envia could embark on a year-long joint research project on voltage fade. If it succeeded—if they solved voltage fade—A123 would buy Envia for $120 million.

  Kumar conveyed the offer to his board, which said that Riley was essentially attempting to steal Envia. If Envia solved voltage fade, the board said, it would not require any relationship with A123—Riley was assuming no risk at all with his offer. Kumar instead should himself go into battle and attempt to solve voltage fade.

  Around this time, Kang was driving to Yellowstone National Park when his cell phone rang. A friend delivered tragic news. Young-Il Jang had died. Just forty-two, he had suffered a freak heart attack in his sleep.

  Absent Young’s intellectual lead, A123 soon became distracted by other matters, including its viability as a co
mpany. Sales were not rising as fast as the founders expected, and a solution to voltage fade dropped from Riley’s agenda.

  Few of the battery guys agreed with Riley and Kang’s assessment. And even if they were right, the market for such powerful batteries was embryonic—it seemed unlikely to soon require battery packs of the size that worried the A123 men. No one in the industry was demanding anything different for the moment either. So no one was in a rush to look deeper at the chemistry.

  But then the industry abruptly changed its mind. GM, Nissan, Ford, and carmakers around the world began to seek batteries that could do much more than boost the mileage of a Prius. They decided that NMC 2.0 stood the best chance of providing such capacity.

  Kang and his young colleague Kevin Gallagher were convinced that voltage fade was an existential problem for NMC 2.0. Unless it could be overcome, the composition was dead as far as use in an automobile. Over several months, they tried but failed to isolate the problem; all they managed to discover were conditions under which it became more severe.

  In October 2010, a big battery conference was to be held in Las Vegas. Until now, only Argonne, Envia, and A123 seemed to be talking about the potential problem of voltage fade. Kang told Gallagher that he ought to disclose it publicly. Gallagher was nervous. Who was he to “rock the boat” on NMC 2.0? But Kang said, “I feel a responsibility. If we know the problem, we have to let it be known so people will work on it. And we should work on it, too.”

  Gallagher delivered the talk. The high-capacity material was still terrific, he said. Except that when you activated the NMC by applying more than 4.5 volts, you triggered fade. He held back from Riley’s apocalyptic forecast but had put the industry on notice that NMC 2.0 had a defect.

  Not long afterward, Kang announced his move to Samsung. Department of Energy staff summoned him to Washington. They wanted to hear more about voltage fade. A few days before his departure to Seoul, Kang sat before six Department of Energy officials with his slide deck. His core message resembled A123’s: NMC 2.0 required a fundamental fix.