Why lithium-ion may rule batteries for a long time to come
The US Department of Energy is launching a major research effort to develop a new generation of lithium-ion batteries largely free of cobalt, a rare and expensive metal delivered through an increasingly troubling supply chain.
Recommended for You
The three-year program, part of a broader effort to accelerate advanced vehicle technologies, could eventually lead to cheaper, longer-lasting consumer gadgets, electric cars, and grid storage.
Materials scientist Gerd Ceder is overseeing one project under the research program at Lawrence Berkeley National Lab, aimed at developing “disordered rock salts” as an alternative material for cathodes, the positive electrode in a rechargeable cell. Typically, the cathodes in lithium-ion batteries require cobalt to create and retain a layered structure in the electrode, which allows lithium ions to easily flow through it. But several years ago, Ceder and his colleagues found that this new class of materials could store more lithium, potentially boosting energy density while avoiding the need for cobalt entirely (see “Disordered materials hold promise for better batteries”).
The Lawrence Berkeley project as well as two at Argonne National Laboratory together received $12.5 million from the DOE's Vehicle Technologies Office.
In an interview with MIT Technology Review, Ceder discussed the challenges to ensure that the new materials work as a “drop-in” alternative for battery manufacturing, the reasons lithium-ion technology will continue to dominate storage for a long time to come—and why it takes so long for any battery advance to reach the marketplace.
This interview has been edited for space and clarity.
What are the next steps in the development of this new class of compounds?
It’s been something like four years now since the initial discovery of the concept, and there’s more than a dozen compounds in this category that have already shown promising features. So that’s the discovery phase, where everyone goes and tries all kinds of different chemistries, like Lewis and Clark exploring.
The next step is we’re going to take some of these materials that look promising and see if we can solve all the small problems that have to be solved before we can actually make a commercial product.
The charge-discharge rate capability has to be good. Cycle life, which sets the lifetime of the battery, has to be improved.
And then people learn to do all kinds of processing tricks and surface treatments, and that’s how batteries get better and better.
But I would say some of these materials are going to go to the next stage. It’s probably one of our best bets these days for higher-energy-density cathode materials.
Why does it take so long to see any lab advances in storage actually make it to the marketplace?
For anything to make it into a commercial product is a long slog, even if you make the discovery faster. It’s just a very long road to materials optimization, testing, customer acceptance, all of these things. To the point that even if I had something that worked perfectly in the lab today, you would probably have a six-to-10-year slog.
Somebody has to scale it up. They have to test it. Then they have to pass it on to a battery maker, cell maker, who will spend two years testing it and then, if they finally like it, they may make a small product out of it—something that goes to a niche market, because they don’t want to take market risks with new products.
A few years ago you said that solid-state batteries are “almost a perfect battery.” What do you think today?
I’m still optimistic that they are a real game-changer. And I think that the ultimate product could be so good, and that justifies, I think, the effort and the push toward it. It could start to look like the ideal battery.
Having said that, of course, in the years since then we and other people have discovered all the issues with it that have to be solved. So I still think it is one of the most promising things for energy storage right now. But there are quite a lot of issues that have to be resolved.
These solid-state electrolytes are often not stable. And nobody’s actually even come up with a great way of manufacturing solid-state batteries.
You can make them in the lab, and companies can even make a prototype, but that just proves you can do it. It doesn’t prove you can do it economically.
A lot of these are engineering challenges, and I have this sort of philosophy that engineering challenges get solved with money. Scientific problems are in a much different category. It’s hard to put a time line on invention.
But I’m confident it will happen.
Grid storage is quite the wild card, in my opinion. In the short term, I just see that all being lithium-ion. The reason is it’s reliable and you can buy it from reliable vendors today.
You can have a philosophical discussion about whether that’s the best form of energy storage for the grid or not. But if you’re PG&E, who are you going to buy from? Startup XYZ, who may not be there in three years, or are you going to buy from LG Chem, CATL, or Samsung?
So I’m not excluding that other technologies can penetrate into the grid, but they should not underestimate the competition from the incumbent. The incumbent is always a powerhouse.
The counterargument you hear is that there are limits to how cheap lithium-ion can get—and it’s not the ideal technology fit for the days, weeks, or even months of storage you might need for the grid (see “The $2.5 trillion reason we can’t rely on batteries to clean up the grid”).
Lithium-ion is pretty cheap already. All these other technologies—while on paper they may have a low cost structure, they still have to get there. Startups need high-value markets, and the grid isn’t a particularly high-value market.
Let’s take an extreme example—let’s look at seasonal storage. How much are you willing to pay for a kilowatt-hour that you move from June to December? The thing has to be unbelievably cheap.
It’s not clear to me that the solutions for those problems will be found in classical electrochemistry storage. Not every problem should be solved with batteries.
AI is here.
Own what happens next at EmTech Digital 2019.