The Powerhouse: Inside the Invention of a Battery to Save the World (24 page)

T
he Hub team moved into a space the size of a regulation basketball court on the first floor of Building 200, across a lawn and a parking lot from Building 205. When you entered, you walked between dark gray cubicles arranged into two rows. All the activity was in the far end of the room, a section separated by a partition and containing a conference table surrounded by soft, black swivel chairs. It seemed intended for important meetings. Chamberlain called it “the War Room.” Butcher paper was taped on a wall, each sheet representing part of the proposal. Two columns were sketched onto four additional sheets that hung on the back wall behind a large video screen. In the left was written, “Days till Submission.” In the right, “Days till Red Team.” The former was fifty-three, the latter twenty-eight.

Chamberlain strode in. He had been delayed by a call with Madia.

“What visionary ideas are we seeking?” he said.

They needed a core rationale for their claim to the Hub. Chamberlain said the question had to be answered in the first five pages of the proposal. The Hub would be won or lost there. If the initial pages were loved, a reviewer would read the next twenty-seven subsections simply to justify the positive impression. But if they failed to do that—if they could not get the first five pages right—all would be in vain.

Gallagher asked, “Are we writing to the FOA or are we reading between the lines?”

“What do you mean?” Chamberlain said.

Gallagher was responding to Madia’s admonition that Argonne had overemphasized basic research. His observation stumbled on the team’s central dynamic: for the six weeks they had been meeting, Chamberlain’s guys had plumbed a single question—what did the Department of Energy
really want?
There was the FOA of course—the department’s own instructions—but surely there was more to it. Chamberlain led this suspicious line of discussion. He would introduce his thoughts with preludes such as, “Our intelligence is telling us that . . .” or “Our guy in the DOE is telling us. . . .” No one knew the identity of these mysterious sources, but Chamberlain’s tone suggested he had the real scoop: the FOA said applied, but it meant research.

Madia had told the group, “Believe the FOA.” In other words, Chamberlain might have his sources, but they were wrong.

So at this moment Gallagher stared intently and spoke slowly, a loose hand gesturing. “Well, until now we’ve been anticipating what we think is between the lines. So are we doing that or writing what the FOA actually says?”

Chamberlain turned to the rest of the group. “In industry,” he said, “I always sought to find out what the customer wanted and fulfill it. In this proposal, I had resisted that—until this review.” He trailed off but the meaning was clear: the old operating principle was dead. They would stop second-guessing.

He flicked a slide on the screen. Its headline was “Goals.” Then it said, “Rapid evolution. Operating prototypes of currently non-existing technology.”

Chamberlain said, “All the work, including research, is aimed at producing functioning prototypes. I mean all work. We aren’t going to produce quantum dots for the sake of making quantum dots. But if we can argue that making quantum dots will allow us down the road to create a functioning prototype, we will.”

Brad Ullrick, an Argonne lawyer, sat near Chamberlain. Neither had actually competed to sit on the proposal team. The pair had swooped in and claimed seats in the Hub. No one else
sought
the positions they now occupied. The two men had unusual skills for a national laboratory—Chamberlain an entrepreneur, Ullrick a legal heavyweight with a background in patent law, both sharp and enterprising. Ullrick was recently divorced. He and his former wife lived a couple of blocks from each other and shared custody of their two young daughters. He was athletic and personable, with a warm smile, and wanted to date. But the bruises of the incompatible marriage were conspicuous in his diffidence. Ullrick seemed to emerge most fully in Chamberlain’s presence. “Jeff’s my best friend,” he said.

A year and a half earlier, Ullrick had suggested to Chamberlain that he—Chamberlain—was destined to be an assistant lab director, a direct deputy to Isaacs. At the time, it seemed an outlandish idea. It was true that Chamberlain had made the unusual jump from the intellectual property unit to Battery Department head. But senior management of the lab as a whole? Now Chamberlain was directing what could be the most consequential project at Argonne in memory and Ullrick’s call seemed less far-fetched.

Ullrick thought it would be healthier if Argonne had more people like them. Perhaps, but it would open up a new set of demands because the lab would then have to figure out how to hold on to these very different character types. If you looked at Chamberlain’s record, he had not stayed with his jobs for long—a few years. No one could know if he would stick around this time. But meanwhile Ullrick planned to ride up with him.

Coghill attempted to orchestrate the team’s methodology. Once, she wanted the group to sit in threes and pass edited pages from one person to the next, providing all the chance to mark up each section. It could take awhile but was extremely effective, she said, resulting in a well-scrubbed document. Instead, a dozen or so team members worked on their sections alone.

One day, Isaacs walked in behind a delivery boy with a stack of pizzas. He told the team that what they were doing was important and made small talk. The effort helped—the Hub team seemed bucked up by Isaacs’s personal interest in the project. Isaacs also noticed a change in Chamberlain, a new decisiveness. It was how he spoke to his people, including to Isaacs himself. Respectful, but clear. “You’ll do this for me,” he would say. Chamberlain had made the necessary transformation.

Someone asked Chamberlain how the proposal was looking. “Crappy,” he said. Madia had told him to expect to finish just three quarters of it by the time of the Red Team review, and that seemed about right. Was he feeling the pressure?

“Yep.”

 • • • 

Madia’s follow-on review came and went. Chamberlain thought he was again “brutal” but also “very helpful” in suggesting specific editing changes. Someone had finally crafted a snappy vision. It was called “five-five-five.” The idea was to achieve five times greater energy density than current state-of-the-art batteries at one fifth the cost, accomplished in the five-year program duration.

George Crabtree, a superconductivity pioneer and member of the National Academy of Sciences, would run the Hub if Argonne won. Crabtree had been added to the Hub team relatively late. The original choice—a university scientist—backed out, which turned out to be a good thing because, while a strong researcher, he was uninspiring and lacked influence beyond the labs. The battery stars would not have followed him and he would have little influence in Washington, which after all would select the winner. That had led Isaacs to Crabtree, a well-liked former Chemistry Division head and one of Argonne’s main ambassadors to the Department of Energy. Crabtree had a gentle way about him and a gift for articulating complex ideas; senior Department of Energy supervisors frequently invited him to contribute to important energy studies and lead public events. What he lacked was any knowledge of batteries. But Isaacs thought he would quickly get up to speed. Crabtree had the scientific stature that the Argonne team needed and so was the right guy. The Red Team agreed.

Thackeray and Amine dismissed five-five-five as fantasy. But the Red Team liked the sound of that, too.

The initial five pages were still not ready. The Red Team “trashed us” on that section, Chamberlain said. It was crucial that Argonne explain how it would meet its goals. Chamberlain said, “We have to rewrite the whole thing.”

Some of the other criticism could be blamed on Chamberlain’s leadership flaws. His inclination to seek consensus was slowing down the work; it was time to take the proposal out of the group and put it into the hands of two or three decisive team members. Generally speaking, Chamberlain said, he needed to “listen to my instincts better.” In one case, the Red Team had suggested he use a chart rather than words to make a point. Within the proposal team a few days earlier, Chamberlain had suggested precisely that approach but had relented against resistance. “I listened to everybody else. Damn it,” he said. Chamberlain said he did not plan to become a dictator but would manage differently. He said, “I am becoming more and more aggressive and confident. Let me put it that way.”

He was experiencing mood swings. Immediately after the review, he concluded that the effort did not stand a chance. It simply could not be brought up to par by the May 31 deadline. Four hours later, he said, “It is looking pretty good. We are going to win.”

Battery guys traveled to Washington, D.C., every May for the Department of Energy’s week-long Annual Merit Review, when the recipients of its approximately $90 million in annual research spending stood before judges to defend their work. This year it was held at a Marriott in Virginia, just across the Potomac River from D.C.

Voltage fade dominated the first day. Peter Faguy, the Department of Energy official, said fade “has become front and center” given the reward to the electric-car industry should NMC 2.0 deliver the promised performance. Amine, Thackeray, and Chris Johnson rose one after another to propose separate solutions to the problem. In the back rows, Gallagher worried about the Hub proposal, due in just over two weeks. So did Jack Vaughey, playing Angry Birds on his smart phone. Both had been elevated to the core writing group, leaving behind their routine duties, and had been working sixty hours a week on the proposal.

That evening, Gallagher and Croy, Thackeray’s protégé, joined Don Hillebrand, the automotive group director, for a dinner of Mexican food. Hillebrand pointed out that politics played a considerable role in large funding decisions and that no one should be surprised if Oak Ridge—a recognized heavyweight—bested Argonne for no reason other than good political or bureaucratic connections. He went on a bit longer, when Gallagher said, “I don’t appreciate what you are saying.” Gallagher said he had been pouring his heart into the project, and here Hillebrand was making light of it. Whatever the case, Hillebrand should shut up. Hillebrand apologized and said he hadn’t meant that at all. The dinner ended a quarter hour later.

Outside, Gallagher texted Chamberlain, who said Hillebrand was a “dick” and had “always been a dick.” Then he texted Tony Burrell, who said Hillebrand was an “asshole” and not to worry about it.

Standing near another door, Hillebrand said, “Either those guys are really nervous or he is an asshole,” meaning Gallagher. “And I suspect it is more of the latter.”

 • • • 

With ten days to go, Chamberlain, Mark Peters, and Holly Coghill huddled in Argonne’s War Room. They had raised $52 million in commitments for outside funding on top of the $120 million Hub award. Their thousand-page proposal would include personal letters from Rahm Emanuel, Chicago’s mayor, and California governor Jerry Brown, along with the governors of Illinois and Michigan, twelve United States senators, and a considerable list of congressmen. They did not expect to produce a perfect document. Nor would they write to the bitter end. One of the final decisions was a superstitious one—they would submit the proposal ahead of time. They would take no risks with the deadline.

On May 25—six days before the due date—Chamberlain hit the send key.

T
wo years earlier, in 2010, GM had found that Kumar’s version of NMC missed a couple of important specs, specifically with regard to a metric called DC resistance. This malady arose when you sought to travel the last twenty or so miles on a one-hundred- or two-hundred-mile battery. The material would put up fierce resistance—the cathode would object to taking in the last of the lithium from the anode. It wanted you instead to plug in the car. The sensation for the motorist was that you suddenly lost power—the vehicle would become exceedingly sluggish. GM said that Kumar had to significantly reduce the DC resistance in his cathode before it could be installed in a commercial car.

But that did not dampen GM’s enthusiasm
for the cathode. The advantages should Kumar succeed were too tantalizing to ignore: his new electrode, more than any that GM had studied, seemed to be on course for a large improvement over the best commercial cells. If he resolved a few hang-ups, his cathode would allow the carmaker to significantly reduce the price of its electrified vehicles and increase the distance they could travel on a single charge. GM engineers continued to counsel caution—given the risks if anything went wrong or if Envia simply fell short of expectations, the carmaker was better off relying on long-standing supply relationships. But Kumar’s ARPA-E announcement—the disclosure that he had broken through existing performance barriers and created a 400-watt-hour-per-kilogram battery—had tipped the discussion. The Envia announcement had been somewhat exaggerated—the start-up suggested that it was achieving 400 watt-hours per kilogram of energy density for hundreds of cycles.
1
In fact, it maintained that milestone for just three cycles before the energy plunged. But this seemed to be overlooked—anyway, it would improve. GM management wanted the Envia battery and ordered its negotiators to close a licensing deal.

About this time, Kumar e-mailed a slide deck to Damon Frisch, GM’s designated intermediary with Envia. It was a blockbuster follow-up to the ARPA-E Summit. Envia, one slide said, had achieved “precise control” of the anode’s silicon-carbon nanostructure. It was starting to scale up production to two-kilogram bulk batches.

The previous summer, GM’s Mark Mathias had told the Argonne battery guys that Kumar’s NMC 2.0 “does not work” as yet. About nine months later, Mathias said he remained on the fence—Envia’s product
might
yet get all the way there or, conversely, could “become the sort of thing that sends people back to the fundamental R&D labs.” Lauckner—who held a decisive vote—said he was impressed with Kumar’s progress. The silicon-carbon anode was “probably by now close to six hundred cycles, which is pretty respectable,” he said. “You will run into a lot of battery company start-ups that will be happy if they can get it to one hundred cycles.” His negotiators continued to hammer out a deal with Kapadia.

Word somehow reached the Department of Energy that Envia was a foreign acquisition target. Honda and Toyota in fact were hounding Kumar for custom NMC 2.0 cathodes and more energetic batteries, and Samsung’s Sun-Ho Kang had told him that he wanted the juiced-up electrode as fast as possible. The DOE sent an alarmed message to Chamberlain: stop Envia from migrating abroad. Chamberlain was furious. Envia itself
was looking for a cash exit. It was not the target of a hostile takeover. And if Envia wished to leave the United States, that was “none of our business.” “I just look at it this way,” he said. “We are a bunch of scientists working for bread. Why are we thinking about industrial policy?”

Chamberlain was either naïve or pretending. If the Department of Energy sought to hold on to Envia, a federal grant recipient with a significant apparent breakthrough, it was acting no differently from Chamberlain himself. Wasn’t he forever speaking of saving America? Wasn’t that what the Hub was all about? The Hub
was
industrial policy. Sitting in his office, Chamberlain conceded that this was true.

 • • • 

Argonne and Envia were not the sole U.S. actors in the race. While they fought for prominence, Elon Musk, the South African chairman of Tesla Motors, became the popular face of electric cars in the country. A lithe, distant, and tall man with furry patches around the perimeter of his face, Musk had earned a fortune by cofounding and selling PayPal. Now, with his exquisitely designed Tesla, he had made electrics cool. To power them, he had in a way endorsed the ExxonMobil forecast: he had snubbed the race for a battery breakthrough and staked his ground on what was available off the shelf. Musk’s bet was that a pure engineering play—a sizzling concentration of high-tech luxury on wheels—could win the market before anyone created a super battery, perhaps long before.

It was a brave, clever, and altogether unpredictable maneuver. Musk, a doyen of Silicon Valley with a bachelor’s degree in physics, was thumbing his nose at scientists: his senior team seemed old school, with a former Toyota executive in charge of manufacturing and a Mazda man as chief designer; only JB Straubel, his chief technical officer, was a standard product of Silicon Valley, with a master’s degree from Stanford and a string of technology jobs. And though they invented no new battery materials, their cars were unlike anyone else’s. They were propelled by “18650s,” cylindrical nickel-cobalt-aluminum batteries with the same general appearance as AAs made for cameras, only larger. The batteries were acquired from Panasonic, some eight thousand in each car. They added about 1,300 pounds of weight to the vehicles but, mounted in the floorboard, they contributed great stability. Using the 18650s meant exchanging development risk and cost for engineering risk: though Musk’s team didn’t have to struggle with physics and invent the better battery, they did have to design a battery
pack
that delivered increasing efficiencies until a model could be sold for roughly $30,000, Musk’s goal for a mass-market electric by the end of the decade. As of now, Teslas sold for double and triple that sum, depending on the model.

For the very reason that his cars did not require a laboratory breakthrough, Musk captured the rapt attention of big incumbents. Toyota, Mercedes-Benz, and Daimler variously bought shares of Tesla and signed deals to acquire its power trains and batteries. It was powerful validation.

Musk said he was not worried by NMC 2.0. Kumar had said he was speaking to Tesla, but Musk remarked, “There are a lot of claims made by battery people.” His team had selected nickel-cobalt-aluminum based on price—it was the cheapest calculated by kilowatt-hour, he said. If he were leading the development of the Chevy Volt, Musk said, he would immediately discard the NMC and switch to his own material.

The Argonne guys disputed the wisdom of Musk’s choice—Gallagher said that nickel-cobalt-aluminum, while having impressive energy density, was also among the most volatile of the main lithium-ion chemistries. It easily caught fire. Musk was courting trouble by putting it in the Tesla.

At ARPA-E, Kumar had unveiled a battery that produced 400 watt-hours per kilogram of energy density. But he would need to seriously improve its cycling performance to make it usable in an electric car—the battery would need to be capable of being charged and discharged 1,000 times. One major hurdle was voltage fade. Thus far, like Musk, he was relying on an engineering solution to stabilize the fade—he was keeping voltage under 4.5 so as not to trigger full-blown fade. You could use the resulting NMC in a working battery. But performance would be much better if it could be reliably activated at 4.6 or more volts—if Envia could deliver NMC 2.0. The start-up could spark a titanic jump if it truly got to the bottom of voltage fade.

Kumar wanted to act fast. It was time to shift from engineering solutions to the deep science. He needed someone familiar with state-of-the-art nanoscale beam-line instruments, technology available only at the national labs. Only such an expert could conceivably deliver those results.

Kumar invited Jason Croy to pay a return visit to Newark.