AI May Have Just Invented The Ultimate Battery | by Will Lockett | Predict | Jun, 2022 | Medium


Artificial intelligence is revolutionising almost every industry on Earth. From optimising farms, to inventing new drugs, to even convincing engineers they are alive, AI is rapidly changing every aspect of our lives. In fact, a team from the University of North Carolina is now using AI to develop the ultimate battery, which will be hugely energy-dense, long-lasting, fast-charging and eco-friendly. But how has this AI managed this incredible feat? When will we get this battery? And what can we do with such outstanding technology? Welcome to the fantastic world of fluoride-ion batteries.

So what are fluoride-ion batteries? Well, functionally, they are identical to the lithium-ion batteries that we use in our phones and cars. This means that the fluoride ions (an ion is an atom with no electrons) pass from the cathode through the electrolyte to the anode to store charge and back the other way to release charge.

However, unlike lithium ions, fluoride ions are far more compact and can pack in together really tightly. This means that, in theory, a fluoride battery can be ten times as energy-dense as today’s lithium-ion batteries. For context, if you replaced the battery pack of a Tesla Model 3 Long Range with one of the same mass while using this theoretical fluoride chemistry, it could go 3,400 miles per charge!

But the advantages of fluoride go far beyond energy density. Fluorine (the elemental version of fluoride) is far more abundant in the Earth’s crust, being the 13th most prevalent element in comparison to lithium’s status as the 33rd most prevalent. What’s more, unlike lithium, fluorine is readily available in common minerals such as fluorite and calcium fluoride. This means that the manufacturing of fluoride-ion batteries could be cheaper and far more eco-friendly than lithium-ion.

So, if they are that good, then what’s the catch? Why aren’t we using this technology? Well, there are a lot of materials that will easily allow lithium ions to pass through them, but there are very few that can do the same for fluoride ions. This makes it incredibly difficult to find materials to build the electrolyte in a fluoride battery that will work, let alone function well. As a result, despite their incredible energy density, most fluoride batteries have been quite crap, lasting only a few charges before massively degrading.It would take decades of arduous calculations and trial and error to find the materials that could unlock this potentially game-changing battery. Or rather, it used to.

This is where J.D. Sundberg and his colleagues from the University of North Carolina come in. You see, they trained an AI to find this wonder material in a matter of weeks, not decades.

Firstly, his team identified 10,000 materials that could potentially do the job. Then, they selected 300 of them at random and calculated their fluoride transporting ability. This was a long and laborious process, given each calculation took a week! Once they had their results, they used them to train an AI. The result was a programme that could accurately calculate any given material’s fluoride ferrying capabilities in only an hour!

The team then ran this programme through the other 9,700 materials in the hope of finding their perfect match, and wouldn’t you believe it, they did.

According to the programme, a zinc and titanium-based material came out on top (ZnTiF⁶). Its amazing storage capacity made it a perfect electrolyte for fluoride batteries, but crucially, it isn’t too expensive and can be synthesised in the lab. What’s more, the raw materials can be mined with far less environmental damage compared to lithium-ion.

In short, this could indicate that Sundberg is on track to develop a high-energy-density, low-cost, and environmentally friendly battery.Such a battery could revolutionise our day-to-day lives and even save the planet from the self-made catastrophe that is climate change.

So what’s next? Well, now S.D. Sundberg needs to run some real-world tests to get this wonder battery up and running. This will most likely include bench testing to find anodes and cathode materials that work best with their new electrolyte material, such as lanthanum nickel oxide (La2NiO4), which has already been shown to work well in fluoride-ion batteries.

From these tests, they can figure out which energy densities, lifecycles, and charge rates are possible. They probably won’t get anywhere near the theoretical limit yet, but the scientists behind this project are confident they can unlock much higher densities than we currently have. If they do, it could change the world. But how?

Firstly, you could buy an affordable electric car with a range north of a thousand miles, which is pretty damn neat. But it could also mean that properly cheap and affordable EVs with ranges per charge in the 250-mile region are possible, allowing everyone to go electric. What’s more, if widely adopted, this technology could significantly reduce the entire industry’s environmental impact. Both of these aspects are vital if we are to transition to a net-zero society and save the world.

But a much more energy-dense battery also means that other EV applications become viable, such as long-haul trucks with high load capacities, ships, and even commercial aeroplanes. All of these industries can’t use current battery tech, given they need compact and lightweight systems to operate correctly, so they are staying with their fossil fuel burning ways and resisting the transition to net-zero. But if Sundberg and his team can create batteries even three times as dense as those we currently have (which is not very dense for fluoride batteries), then the EV revolution can start to happen in these carbon-belching industries, yet again allowing us to stop climate change and save the world.

So that is how AI may have just invented the ultimate battery. By quickly identifying materials worth testing, we can speed up the development process of next-gen batteries that can help the entire world go carbon neutral without breaking the bank or causing another ecological disaster. It may still be a decade or two before this technology sees the light of day, and even longer before it is produced cheaply and at scale, but with technology like this just over the horizon, it can give us hope that we can fight climate change and save the world. Let’s pray that Sundberg and the other brilliant scientists working on fluoride-ion batteries can bring this incredible technology to fruition before it is too late.