So Daniel sent us this one. He’s asking us to dive into batteries. Not the most glamorous subject, he admits, but they’re in everything. He wants to know why lithium-ion became the dominant, ubiquitous chemistry. What are the main advances pushing the bounds of capacity, recharge efficiency, and energy density? And then, what’s on the frontier? Solid-state cells moving from lab to limited production, silicon-anode variants, sodium-ion as a cheaper alternative. The gist is, this frontier is closer than ever to reshaping everything from smartphones to EVs. So, where do we even start with this, Herman? The periodic table?
We kind of have to start there, yes. Because the dominance of lithium-ion isn't an accident of marketing. It's a fundamental consequence of physics and chemistry. Lithium is the third lightest element on the periodic table, and it's also the most electropositive. That means it really, really wants to give up its outer electron. In battery terms, that translates to a very high electrochemical potential. You get a lot of voltage per cell, which is the electrical "pressure," and you get it from a very light material. That combination is the holy grail for energy density—how much energy you can store per kilogram.
So it’s not that we collectively decided lithium was cool. It’s that the universe handed us the best possible lightweight ingredient for storing charge.
Right. The Wright brothers problem again. The bottleneck wasn't the idea of flight; it was materials. For batteries, the periodic table is the constraint. You can't invent a better element than lithium for this purpose. The challenge has always been building a stable, safe, reversible system around this very reactive, very energetic element. That's what the "ion" part of lithium-ion is about. We're not using metallic lithium, which is dangerously unstable. We're shuttling lithium ions back and forth between two electrodes.
Let’s make that concrete for a second. When you say “dangerously unstable” for metallic lithium, what does that look like in practice? Is this a lab curiosity, or have we seen it?
Oh, it’s very real. The first attempts at rechargeable lithium batteries in the 70s and 80s used lithium metal anodes. They were a fire hazard. The lithium would form dendrites—those tiny, needle-like structures—on charging, eventually piercing the separator and causing a short circuit, leading to thermal runaway. It was so problematic it shelved the technology for consumer use. There’s a famous case in the 1980s with Moli Energy, a Canadian company. They launched a lithium-metal battery for cell phones, had to recall it after multiple fires, and the company went bankrupt. That failure directly paved the way for the “ion” approach—shuttling ions between stable hosts, which Sony perfected.
So the "ion" is the compromise. We trade the ultimate theoretical performance of pure lithium metal for stability and cycle life.
And that compromise defined the last thirty years. The modern lithium-ion battery, with a graphite anode and a lithium cobalt oxide cathode, was commercialized by Sony in nineteen ninety-one. It won because it worked demonstrably better than anything else for portable electronics. Nickel-cadmium batteries had memory effect, they were toxic. Nickel-metal hydride was better, but still lower energy density. Lithium-ion was a leap. It had higher voltage, higher energy density, no memory effect. The trade-offs were cost and, initially, safety concerns that required sophisticated battery management systems.
Which, to be fair, we’ve seen play out. The Note Seven incidents, the occasional hoverboard fire. The chemistry is inherently energetic. So the entire engineering discipline around it has been about containing that potential, literally.
That's the ongoing story. The core chemistry has been remarkably stable for over thirty years. Almost all the gains since the nineties have come from incremental engineering. Tweaking the cathode materials—moving from lithium cobalt oxide to lithium iron phosphate for safety and cost, or to nickel-manganese-cobalt blends for higher energy density. Improving the electrolyte additives to form better solid-electrolyte interphase layers. Thinner, more precise separators. The gains have been a steady two to three percent per year in energy density, which compounds dramatically.
Can you give us a sense of that compounding? Like, from my first iPod to my current phone, what did that look like?
Sure. A state-of-the-art lithium-ion cell in the early 2000s had an energy density of about 100 watt-hours per kilogram. Today’s best consumer cells are pushing 300 Wh/kg. That’s triple, from those tiny year-over-year improvements. Your modern smartphone battery holds about three times the energy of one from twenty years ago, in the same volume, and it charges ten times faster. That’s the power of incrementalism. But we’re starting to bump against material limits in the current architecture.
So when we talk about pushing the bounds now, we're not talking about a new two-percent tweak. We're talking about step changes. What's the biggest lever being pulled right now?
The anode. For decades, we've used graphite. It's stable, it's cheap, it works. But it has a theoretical capacity limit for how many lithium ions it can hold. Silicon, on the other hand, can hold about ten times more lithium per atom. The problem is that when silicon absorbs all those lithium ions, it swells. Dramatically. Up to three hundred percent volume expansion. It pulverizes itself over charge-discharge cycles.
So the advance isn't discovering silicon is good. It's figuring out how to stop it from exploding like a overfilled pastry.
Precisely. I mean, yes. The breakthroughs have been in nano-engineering. Instead of a solid silicon block, you create silicon nanostructures—nanoparticles, nanowires, porous silicon frameworks. These structures have room to expand internally without fracturing the overall electrode matrix. Think of it like replacing a solid concrete block with a sponge. The sponge can absorb water and expand within its own pores without shattering.
That’s a great analogy. So you’re creating engineered empty space at a microscopic level.
The other approach is silicon-dominant composites, where you blend silicon with graphite and special binders. We're seeing this enter the market now. Tesla's four-six-eight-whatever battery day talked about their silicon anode design using an elastic polymer binder and a conductive coating. Several Chinese EV manufacturers are shipping cars with batteries using silicon oxide or silicon-carbon composite anodes. The gain is real—fifteen to twenty percent increase in energy density at the cell level, sometimes more.
And that translates directly to either a lighter car with the same range, or a longer range for the same weight.
For phones, it could mean a thinner device with the same battery life, or a device that lasts two days. It's a material science problem that's yielding to engineering. The other major lever is the cathode. There's a push towards what are called "high-nickel" N-M-C chemistries. Reducing the cobalt content, which is expensive and has ethical sourcing concerns, and increasing the nickel content, which boosts energy density. But high-nickel cathodes are less stable, more prone to thermal runaway. So again, it's an engineering challenge: coating the cathode particles, stabilizing the electrolyte, improving thermal management.
Let's talk about that thermal management bit. Because it feels like the unsung hero. The battery pack in a modern EV isn't just a box of cells; it's a climate-controlled apartment complex.
It absolutely is. The battery management system is a masterpiece of conservative engineering. It's constantly monitoring the voltage, temperature, and health of hundreds or thousands of individual cells, balancing them, keeping them in a perfect Goldilocks zone. This is where a lot of the "recharge efficiency" gains come from. Minimizing losses as heat. Liquid cooling plates, sophisticated algorithms that precondition the battery for fast charging. The difference between a car that can sustain a high charging rate and one that throttles after ten minutes is often in the thermal system, not just the cell chemistry.
Can you give an example of how sophisticated this gets? Is it just a liquid loop, or is there more to it?
Oh, it’s wild. In some premium EVs, the thermal system is integrated with the car’s HVAC. On a hot day, it will actively cool the battery pack, even if the car is parked and plugged in, to prepare it for fast charging. On a cold day, it will warm the battery using a heat pump or resistive heater to get it into its optimal temperature window before you even start driving or plug into a charger. Some systems even use phase-change materials that absorb heat as they melt. It’s a full-time, climate-controlled environment for these sensitive chemical reactors.
That reminds me of a fun fact I came across. Some high-performance EVs, like certain Porsches, have a "battery conditioning" mode you can activate from your phone app before a track day. It’s not just about charging; it’s about pre-cooling the pack to handle sustained high-power output without overheating. It treats the battery like an athlete warming up.
That’s a perfect example. It’s systems thinking. The cell is just one component. The real performance is in the system built around it. And that system is getting incredibly smart.
Which brings us to the poster child for the next generation: solid-state batteries. Every other headline claims they're a year away. What's the real promise, and what's the actual holdup?
The promise is transformative. Replace the liquid organic electrolyte—which is flammable and the source of most battery fires—with a solid ceramic or polymer electrolyte. This does several things simultaneously. It dramatically improves safety, as there's no flammable liquid to leak or ignite. It potentially allows for the use of a pure lithium metal anode, because the solid electrolyte can suppress the growth of lithium dendrites, those needle-like structures that short out liquid cells.
Dendrites. The bane of metallic lithium anodes.
Right. And if you can use a lithium metal anode, the energy density jump is enormous. We're talking double or more the energy density of today's best lithium-ion cells. You also get faster charging, because ion transport can be more efficient in some solid materials. The holdup is not the science; we've had lab prototypes for decades. The holdup is manufacturing. Making thin, flawless, durable solid electrolyte layers at scale, at a reasonable cost, is brutally hard. These ceramics can be brittle. Ensuring perfect, low-resistance contact between the solid electrolyte and the electrodes across thousands of cycles is a materials engineering nightmare.
So who's actually close? Not "close" in press release terms, but in "we have a pilot line and are shipping limited batches to automakers for testing" terms?
Toyota has been the most conservative and also, paradoxically, the most vocal. They've been working on solid-state for years and have recently shown a prototype with a claimed range of over seven hundred miles. They're targeting limited production by twenty twenty-seven or twenty twenty-eight. A company called QuantumScape, which went public via SPAC, has a partnership with Volkswagen. They're focusing on a ceramic separator and lithium metal anode. Their latest data shows promising cycle life, but scaling is the question. Then there are Chinese players like WeLion, which is reportedly already supplying small volumes to Nio for test vehicles. The consensus is we'll see solid-state first in premium EVs, maybe in drones or aviation, where the weight savings justify a huge cost premium. It won't be in your phone for a long, long time.
Because the cost per kilowatt-hour will be astronomical initially.
It's a new materials supply chain from the ground up. Now, there's an intermediate step that's getting less press but is arguably more imminent: semi-solid state or gel polymer electrolytes. These are thicker, quasi-solid electrolytes that offer some of the safety benefits without the insane manufacturing challenges of brittle ceramics. Companies like S-E-E-O, now defunct, worked on this, and some Chinese firms are pursuing it. It's a pragmatic stepping stone.
Okay, so we have silicon anodes squeezing more from the current paradigm, and solid-state as the potential revolution. But Daniel also mentioned sodium-ion. That sounds like a step backward on the periodic table. Why?
Because sodium is everywhere. It's in table salt. It's absurdly cheap and abundant. Lithium deposits are concentrated in a few places—Chile, Australia, China—and mining and refining it is environmentally intensive. Sodium-ion batteries work on a similar "rocking chair" principle, shuttling sodium ions instead of lithium. The voltage is lower, so the energy density is lower, maybe comparable to early lithium iron phosphate cells.
So not for your premium EV where you need maximum range.
Right. But think about stationary storage for the grid. Weight and volume don't matter as much in a shipping container-sized battery farm. Cost per kilowatt-hour is king. Or think about low-cost, short-range urban EVs in markets like India or Southeast Asia. Or backup power systems. Sodium-ion is about democratizing storage, not maximizing performance. The Chinese battery giant C-A-T-L is already in mass production of sodium-ion cells, and they're going into some microcars. The technology is here, it's cheap, and it fills a crucial niche in the overall ecosystem. It's not a lithium-ion killer; it's a lead-acid battery and a low-end lithium-ion killer.
Fun fact moment: I was reading that because sodium ions are larger than lithium ions, the cathode and anode materials are often completely different. They can use things like Prussian blue analogs or layered metal oxides that are cheaper and more abundant. It’s a whole different materials playbook.
That’s a great point. It’s not just a drop-in replacement. It’s a parallel technology branch. And that diversity is a strength. It means we’re not putting all our eggs in one mineral basket. There's another interesting angle: sodium-ion batteries often perform better at low temperatures and have a wider safe operating range than standard lithium-ion. So for certain rugged, non-weight-sensitive applications, that's a real advantage.
That's a crucial point. The future isn't one winner. It's a portfolio. Silicon-anode lithium-ion for performance applications. Solid-state for ultra-performance where cost is no object, eventually. Sodium-ion for mass-market, cost-sensitive storage. Lithium iron phosphate for safety-critical, durable applications like buses or grid storage.
You've got it. The battery ecosystem is diversifying. For thirty years, it was basically one chemistry with minor variations. Now we're seeing speciation. And this is being driven by the sheer scale of demand. The automotive industry is orders of magnitude larger than the consumer electronics industry that birthed lithium-ion. That volume justifies investing in multiple technological paths.
Let's talk about that demand for a second. The environmental angle is the constant critique. "Your green EV is powered by dirty mining." What's the realistic path for improving that? Is it just better mining practices, or is there a chemistry shift that alleviates it?
It's both. On the mining side, there's a huge push for direct lithium extraction from brine, which uses less water and land than traditional evaporation ponds. Recycling is becoming economically imperative as the first wave of EV batteries reaches end-of-life. A robust recycling loop can eventually supply a significant portion of lithium, cobalt, and nickel. But chemistry shifts are the bigger lever. Cobalt is the biggest ethical headache. The move to lithium iron phosphate, which uses no cobalt or nickel, is massive. Tesla is using it in their standard-range vehicles. Most Chinese EVs use it. It's cheaper, safer, and more durable, if slightly lower energy density. That's a huge environmental and ethical win. The high-nickel, low-cobalt cathodes are another step. And sodium-ion, of course, avoids the critical minerals issue almost entirely.
So the market is already self-correcting towards less problematic materials, driven by cost and supply chain security as much as ethics.
The geopolitical dimension is unavoidable. China dominates the battery supply chain, from processing minerals to manufacturing cells. Europe and America are trying to build their own capacity with laws like the Inflation Reduction Act, which ties EV tax credits to North American assembly and mineral sourcing. This is accelerating investment in alternative chemistries that use locally available materials. Sodium-ion is attractive to Europe because they have no lithium but plenty of salt. Lithium iron phosphate is attractive because the iron and phosphate supply chains are more globally distributed.
Shifting gears slightly. Fast charging. It feels like the other arms race. We went from "trickle charge overnight" to "twenty minutes for eighty percent." What's the physical limit? And what breaks when you push it too hard?
The limit is largely about ion transport speed and heat. Pushing lithium ions into the graphite anode too quickly causes them to plate as metallic lithium on the surface instead of intercalating neatly between the graphene layers. This is lithium plating. It's irreversible capacity loss, and it can create dendrites. The heat generated can degrade the electrolyte and cathode. So the advances are in cell design to reduce internal resistance. Tabless designs, like Tesla's four-six-eight, shorten the path electrons have to travel. Better thermal pathways. And, crucially, battery management that dynamically controls the charge rate based on the cell's temperature and state of charge. The best systems today will charge blisteringly fast from a low state of charge up to about fifty or sixty percent, then taper off dramatically to protect the battery.
So the advertised peak rate is almost a marketing number. The sustained rate is what matters.
Yes. And the next step is to design cells specifically for fast charging from the ground up. Different anode materials, like lithium titanate, can charge incredibly fast but sacrifice energy density. It's always a trade-off. For the average consumer, the real win isn't just faster charging, but more ubiquitous chargers. A fifteen-minute charge is only useful if you can find a charger when you need it.
What about the charging infrastructure itself? Is the grid ready for a parking lot full of cars all pulling 350 kilowatts simultaneously?
That’s the trillion-dollar question. No, not currently. A single ultra-fast charger can draw as much power as a small neighborhood. The grid upgrades needed are monumental. This is where smart charging and vehicle-to-grid technology come in—where your car battery can actually supply power back to the grid during peak demand. It turns the EV fleet from a grid liability into a potential asset. But that’s a whole other weird prompt.
Noted for the future. But on the vehicle-to-grid point—is that a real near-term thing, or is it still theoretical? I’ve heard about pilots, but what’s the actual holdup?
The holdup is a combination of technology standards, economics, and consumer behavior. Technically, it requires bi-directional chargers and cars equipped to handle it. Economically, you need utility companies to create attractive payment structures for selling power back. And for consumers, there’s the worry about extra battery degradation. But it’s moving past theory. Ford’s F-150 Lightning can power a house in a blackout. In Japan, Nissan has been running vehicle-to-grid pilots for years. In the UK, there are trials where EV owners get paid to let the grid use their car’s battery during evening peak times. It’s coming, but it’s a slow rollout because it involves so many stakeholders.
Practical takeaway time. For someone buying a phone or an EV today, or in the next year, what should they look for? What's marketing fluff and what's a real advance?
For phones, the term "silicon-carbon anode" or "next-generation anode" is usually meaningful. It might get you an extra hour or two. "Graphene battery" is almost always fluff—it usually means they used a trace amount of graphene as a conductive additive, not a fundamental change. For EVs, the chemistry is key. Lithium iron phosphate means great longevity, safety, and lower cost, but slightly less range in cold weather. Nickel-rich chemistries mean higher energy density for longer range. Ask about the charge curve, not just the peak rate. A car that holds a high rate from twenty to eighty percent is better than one with a high peak that plummets after ten minutes. And ignore any claim about solid-state batteries in a production vehicle you can buy today. It's not happening yet.
What about second-life applications? My EV battery degrades to eighty percent capacity after ten years. It's useless for the car, but still a huge battery.
That's a massive emerging market. Stationary storage for homes or the grid is perfect. The cycles are slower, the environment is controlled. Companies are already setting up businesses to test, repackage, and resell used EV packs. It improves the overall economics and environmental footprint dramatically. Your old car battery might power your house for a decade. There’s a great pilot project in the UK where used Nissan Leaf batteries are being used as backup power for streetlights and EV chargers.
That’s a perfect circular economy example. It also takes some of the sting out of that initial battery cost if you know it has a valuable second act.
It changes the total cost of ownership math. And from a grid perspective, these second-life packs are incredibly cheap storage capacity. They don’t need to be as energy-dense or lightweight, they just need to be reliable and safe.
One forward-looking thought. We've talked about the materials and the engineering. What about the fundamental physics? Is there a theoretical ceiling for electrochemical batteries that we're approaching? Do we eventually hit a wall and have to look to something else entirely, like hydrogen fuel cells or supercapacitors?
For lithium-ion and its direct descendants, yes, there is a ceiling dictated by the energy stored in chemical bonds. We're probably within a factor of two or three of that ultimate limit for practical systems. Solid-state with lithium metal might get us close to that ceiling. Beyond that, you're looking at different energy carriers. Hydrogen has its own massive storage and efficiency problems. Supercapacitors have fantastic power density but terrible energy density—they're for bursts of power, not storage. The most interesting long-term bets might be on things like lithium-air or lithium-sulfur batteries, which have staggeringly high theoretical energy densities, but have been stuck in the lab for decades with fundamental stability issues.
What’s the dream with lithium-air? That sounds sci-fi.
It is. The idea is to use oxygen from the air as the cathode material, reacting with lithium. Its theoretical energy density rivals gasoline. But the practical problems are immense: moisture from the air poisons the cell, the reaction products clog the cathode, and efficiency is terrible. It’s a great example of a “moonshot” chemistry that, if solved, would change everything, but it’s perpetually a decade or two away. The incremental engineering path for lithium-ion and its cousins has at least another decade or two of meaningful gains. The revolution, if it comes, will be from solving one of those fundamental chemistry problems we've been staring at for fifty years.
So the weird prompt about batteries ends up being a story about incremental human ingenuity wrestling with the periodic table. We take the best element God gave us, and we spend decades figuring out how to build a safe, reliable, and increasingly cheap box around it. And just when that box seems optimized, we start figuring out how to rebuild it from entirely new materials.
That's a perfect summary. It's not a story of sudden, disruptive magic. It's a story of compounding two-percent improvements in materials science, manufacturing precision, and systems engineering, punctuated by the occasional step-change when a long-standing problem, like silicon swelling, finally cracks. And it's happening faster now because the economic incentive is planetary in scale. We’re not just powering phones anymore; we’re trying to power the transition of the entire transportation and energy grid. That’s a powerful motivator. It’s worth remembering that the lithium-ion battery itself was once a lab curiosity that took twenty years to commercialize. The same could be true for the technologies we’re just starting to see in pilot lines today.
I think that's a wrap on the power cell. Thanks, as always, to our producer, Hilbert Flumingtop, for keeping the electrons flowing. And thanks to Modal for providing the serverless G-P-U platform that powers our pipeline. If you want more deep dives into the unglamorous tech that makes the modern world work, leave us a review wherever you listen. It helps more people find the show. For Corn and Herman, this has been My Weird Prompts.
Walk slowly and carry a big battery.
