Eight Years After the Collapse of Envia, Its Former CTO Has a New Battery
A supercautious Sujeet Kumar is attracting positive reviews for his silicon anode
In 2015, I published a book called The Powerhouse, which followed a group of researchers at Argonne National Laboratory as they attempted to create a super battery. Two members of this group had invented NMC, which later became the industry’s go-to battery chemistry, used by everyone except Tesla. Now, it was trying to move to NMC 2.0, a turbo-charged advance of the formulation.
But on the sidelines, a Silicon Valley startup called Envia Systems was racing to accomplish the same thing. Envia had licensed Argonne’s invention, and gotten GM intensely interested in its version of the chemistry. So interested that GM cut a deal to use Envia’s NMC 2.0 in its first new pure electric vehicle — the Chevy Bolt. It was a gargantuan vote of confidence in Envia, and stood to make its executives and investors wealthy. One of them was Envia’s chief technology officer, a friendly and soft-spoken materials scientist named Sujeet Kumar, whom I came to know over a two-year period of reporting.
Only, just months before the book was due, I learned something stunning — the assertions that Kumar and Envia had been making publicly and to GM about their battery were exaggerated. Envia’s NMC 2.0 had not achieved the astounding metrics it was claiming, and there was no indication that it could any time soon. In 2013, GM canceled the deal, and Envia became a modern poster child for Thomas Edison’s famous 1883 dictum about advanced new batteries — that they are “a mechanism for swindling by stocking companies.”
Eight years later, Kumar is back. His surprising new venture is a startup called Zenlabs, and battery researchers say they are impressed with performance data from his silicon-based anode. “The data does look good,” said Venkat Srinivasan, director of the Argonne Collaborative Center for Energy Storage Science. “Silicon does look closer than it has ever been. And good to see Sujeet back in action.”
The technology that Kumar is attempting to commercialize — the silicon anode — is one of the most sought-after breakthroughs in advanced batteries for electric vehicles. The reason is its electrochemical superiority compared with conventional graphite anodes: Silicon can store up to 4.4 lithium atoms for every silicon atom. But graphite stores only one lithium atom for every six carbon atoms.
The problem is that, when lithium moves into the anode during the charge-recharge cycle, the silicon swells and short circuits the battery. Researchers have managed to sprinkle only a few crumbs of silicon into an anode to get a bit of extra energy.
In the last couple of years, a few companies have claimed they are very close to solving the swelling problem. Kumar is one of them. When I emailed him, it was our first contact since the Envia debacle. He agreed to respond to technical questions by email but did not want to talk about the past, meaning Envia. That seemed fair.
This time, Kumar is playing it extremely cautiously. He said that every year, he has sent his material for independent, third-party testing at one of the federal government’s national laboratories. This way, it is not Zenlabs or he who is claiming any performance metrics, but an independent source.
The data does not say how much silicon is in Zenlabs’ anode, but suggests it is more than half of the electrode. As a result, one type of his battery cells has achieved specific energy density of 315 watt-hours per kilogram through 1,000 charge-discharge cycles, according to the data. Kumar said these cells charge to 90% capacity in 15 minutes and 80% in 10 minutes. Though that is only lab data and not scaled up, a stage at which many problems can and do occur, it is still very high specific energy density and very fast charging. (By comparison, Tesla’s state-of-the-art batteries are thought to achieve around 290 watt-hours per kilogram.) To achieve this performance, Kumar twinned his anode with a cathode made of NMC 622, which is the industry standard and not experimental.
“Our mission is ‘fast charge every charge,’” Kumar told me.
Kumar described a second set of cells that had achieved 350 watt-hours per kilogram using NMC 811, a cutting-edge variety of the chemistry being used by a few automakers, especially in China. The higher specific energy density means the vehicle can go further on a single charge, and that the battery will cost less.
It is a highly competitive time in batteries. Companies such as QuantumScape and Solid Power have claimed equally impressive or better performance data, and say they plan to have futuristic metallic lithium batteries ready for commercialization in the middle of the decade. Then there is Sila Nanotechnologies, which is also producing a silicon anode and is thus a direct competitor of Kumar’s.
Kumar said his anode is currently being tested by automakers. Like QuantumScape, he expects to be commercially ready by mid-decade. His business model is to license the anode rather than make it himself.
There is bound to be skepticism in parts of the battery community. When Envia blew up, however, I concluded — and wrote — not that Kumar was up to no good, but that he had gotten carried away. When GM came sniffing around, he talked up his lab’s stuff. Like inventors of all types, he desperately wanted his work to succeed, and was wholly convinced that it would. Only, in this case, before he knew it, he was actually promising GM that his team would succeed, according to a specific timetable.
That was way too aggressive; to this day, no one has accomplished what Kumar was promising — a 400 watt-hour per kilogram battery. Now, though, Zenlabs is among a clutch of companies that actually do look to be on the cusp of delivering just that.
[Here are links thoughtfully assembled by Mobilist reader Sam DeBrule]