Venkat Srinivasan is the director of the Argonne Collaborative Center for Energy Storage Science (ACCESS) and deputy director of JCESR, which was formed in 2012 as a fundamental research and development arm of Argonne’s portfolio. JCESR (pronounced jay cesear) is a multi-institutional partnership focused on beyond lithium-ion technology, the next generation for transportation and grid.
In this interview, BEST delves into the topics of recycling, long-duration energy and next-generation energy storage solutions, as well as exploring the methods being utilised to bring those technologies to commercialisation.
BEST: How closely do the trends in the battery industry inform your own research and how do you choose which of the different technologies to study?
Srinivasan: We are constantly watching and seeing what’s happening in the world. A great example of this is the growing need that we’ve seen for the supply of materials.
There are people at Argonne that have been talking about recycling for a long time. We’ve been paying attention to it, and some years ago we realised that the trend in terms of adoption was changing, to the point where we knew the time was right to begin research in this area. That’s when we created a strategic area of recycling. Out of that, we now have a centre led by Argonne, that’s called ReCell Center, focused on recycling.
Now it’s front and centre for us. So that’s a great example of how we like to watch and see what’s happening in the marketplace because it tells us where the biggest needs are. This allows us to adapt our R&D to make sure that we are able to move toward the areas where industry wants to go so that we can provide them with solutions when they get there.
It’s the same reason we continue to have a diversity of battery chemistries that we focus on. I remember some years ago working on lithium metal batteries and people from industry saying ‘well, this seems so futuristic’. Well, not anymore.
It’s always been part of the portfolio of the battery community to work on solid-state batteries, but now it’s become front and centre. So, as we watch and see where industry is going, we like to also push them toward where they should be going.
So I would say, it’s a bit of a push and pull; we are watching and seeing where they’re going, to make sure they’re responsive; but we are also trying to push them into new chemistries so that they can think about things in a more diverse fashion— you know, a kind of global arc, if you want to think of it that way.
We all used to work on HEVs— hybrid electric vehicle— batteries for high power with less energy. We then started going into PHEVs where we increased energy some more, but don’t want to lose the power because that’s also important.
And now it’s all about energy, as we’ve seen the marketplace for electric cars change, we’ve also adapted R&D so we are able to pay attention to that.
And it’s the same for the grid. There’s all this focus today on what we call long-duration storage, but that’s another trend that we watched in the early days, realised the big need and adapted the R&D so that we could start going toward long-duration storage. Very early on we realised that it was going to be an important market, and so we quickly moved into that area, as part of JCESR.
Is that you moving into flow batteries and long-duration technologies?
That seems to be the new approach for the lead industry; instead of thinking, ‘it has to be lead’, it’s ‘lead has a place next to lithium-ion, which has a place with next-generation technology’. It’s all part of a global solution.
Exactly. Talking about lead-acid batteries, we have a large project with the ALABC (Advanced lead-acid Battery Consortium), which is a consortium of lead-acid battery companies. And that’s a great example of how we are trying to work with industry. The lead industry had a need. They were trying to understand how to bring the latest scientific tools to bear, to better understand lead-acid batteries.
And they recognise Argonne has some deep capabilities in this area, including the Advanced Photon Source— which is basically an extremely bright light source that enables us to look inside the battery, to understand how it is functioning while it is charging and discharging, without even opening it up.
So we use tools and techniques like this to help the industry understand what’s happening on the inside of the battery and, frankly, our industry collaborators are amazed at what we can see.
And again, it’s us at the laboratory trying to help this consortium of companies understand the fundamentals from which, hopefully, they will be able to use the understanding to start coming up with solutions that can have an impact in the marketplace.
What will be the biggest change to the battery industry in the next 10 years?
If you ask me in the next 10 to 20 years ‘what should we be?’ we should be an industry where we think about the whole cradle-to-grave aspect of batteries, and ask how can we make it into something that is completely sustainable, where every material— how it’s processed, how it’s made— is environmentally friendly and uses renewable sources and ultimately when it’s used in the device, has a tremendous impact. That’s really what I think the community needs to think about.
Number two, I think at the end of the day you’re going to have batteries everywhere, it’s going to become like microprocessors. Everything is going to have a battery.
How do you make it ubiquitous, so that you’re harvesting energy everywhere, and using it everywhere so that we can start to view this world as one where that becomes just part of our lives, and it’s all renewable?
How has/is technology, such as computer simulations and AI, changed the way Argonne works, and in which ways do you see this technology benefitting the battery industry in the long term?
I’m a mathematical modeller, so this is near and dear to my heart. Math is becoming more and more important in the battery industry. We all use the word co-design. What that means is, you have to think about everything from what is the ultimate use: so that could be energy density, power density or cycle life, all the way back to what the material is going to do at the atomic and molecular level.
So we use mathematics to tell us how we should design materials to get performance metrics like energy density, power density, and cycle life.
The second place where math comes in is discovery, and this is where the AI (artificial intelligence) machine learning (ML) part comes in in a big way.
The discovery of materials is extremely important right now, in terms of cathodes, electrolytes, separators; we’re always discovering something in the battery world, and the old way of discovery used to be what I call ‘cook and look’, where you go cook something up, and then you go look and see if it works as a battery.
And, that’s fine, except it’s slow. In the last decade, we’ve seen the emergence of a better way to do it, a faster way to do it, using computation. Today we can calculate all sorts of properties of battery materials that a decade ago we didn’t know how to calculate:
- How much lithium can a cathode hold?
- What voltage does lithium go in and out?
- How fast does lithium move?
- What kind of reactions will that have when it reaches an interface?
- Will it be the reaction that you want, or does it do all sorts of other bad things that you don’t want?
All of these questions are things that we can answer on the computer.
And what we’re learning now is that because computing has become inexpensive— meaning supercomputers are everywhere now— your desktop has become so powerful that we can do many, many, many 1,000s and 1,000s of simulations very quickly.
So, now we have a tremendous amount of data, we need to ask ourselves how we’re going to analyse the data. And so this is where machine learning has come in— to help us understand exactly how we are going to take these models that are producing tonnes of information and, using computational tools, to understand trends and therefore start to extrapolate and see what kind of functionality can be put into battery materials.
Ultimately, this whole idea of AI and ML is: it is going to be used everywhere. One great example of that, and one of the big things in the battery world, has always been, ‘can I take one year of data of cycle life from testing and predict if the batteries will last 10 years, 15 years, 20 years or even more?’ And this has been a constant challenge.
Today, we are learning that we can use machine learning tools, where we can take six months of data, or one year of data, and start to analyse every part of this recharging and discharging curve, and start to ask, ‘Are we seeing features that are early indicators of cycle-life limitations?’
Some of this is just not possible without computers, and without artificial intelligence, because there’s just too much information coming in; but with the advent of AI and ML we can now ask those questions. So that’s become a very important part of what we are thinking of now, ‘How do you start to use AI and ML for things like battery life prediction, for discovery; and how do you start thinking about the synthesis of new materials using AI and ML?’
What percentage of Argonne’s work is testing compared with R&D? Is it 50/50?
We have a large-scale testing facility, which is really an important part of what we do; standard testing for both transportation and grid sectors, where we put the battery through the paces; we do accelerate testing and then we do price prediction. It’s probably 10% of our portfolio, but it’s a very important part.
In 2009, Argonne started a programme to improve lithium-ion. What progress has been made and is it still a focus for the laboratory?
Our lithium-ion work actually started in the early 1990s. Today we continue to push the boundaries on lithium-ion. I won’t go into too many details, but we are really focused on cathodes that have low nickel and low cobalt because that’s an important challenge. We are working on changing the anode to go towards silicon, and ultimately to lithium metal and solid-state batteries.
We are working on removing the flammability of the lithium battery electrolytes, trying to get them to be more stable at high voltage and low voltage, which also makes them less reactive. We have a big programme on fast-charging lithium-ion batteries. We have a goal of charging batteries for less than 10 minutes. And we have a big programme, as I was mentioning, on recycling lithium-ion batteries. So we absolutely continue to look at all of these aspects, and we will probably continue to look at many of these aspects going into the near future.
On JCSER’s website, it mentions specifically flow batteries, and solid-state lithium-ion, which are the ones in the public eye at the moment. But, are you working on chemistries beyond those?
There are three big focus areas in JCESR right now. One of them is on solid-state, trying to understand what makes a solid-state electrolyte work so we can ultimately design new ones that perform better, and therefore can we design new ones to remove some of the fundamental bottlenecks with existing ones. That’s one big area.
For the second big area, we are focused on flow batteries, in particular, flow batteries using synthetic organics as the reactants. So we think there is a world out there, where we can use synthetic molecules, meaning molecules that are made in the lab, that can serve as active materials for flow batteries. We’ve been exploring this and we are putting various functionalities into these molecules.
And the third big area is what we call multivalent batteries. So this is magnesium, zinc, calcium— where we are very interested in inventing new electrolytes, new capital materials and new ways of protecting the interface between the anode and the electrolyte, and the cathode and electrolyte; so that we can start to provide this diversity of battery chemistries beyond the lithium-ion world.
- Read the first part of the interview below
- Read the story Argonne, from how it all began to how it continues to push the boundaries of battery development today HERE
Art of the possible: how Argonne National Laboratory made battery history
