Beyond Lithium-Ion: Sodium-Ion and Solid-State Batteries

Most coverage treats sodium-ion and solid-state batteries as two contestants in the same race to replace lithium. They aren't. Solid-state batteries, in nearly every serious program, are still lithium batteries, and many aim to use more lithium than today's cells, as a lithium-metal anode. Sodium-ion batteries are the ones actually leaving lithium behind, and they do it by giving up energy density, not by gaining it.

So "beyond lithium" is half right. One technology moves beyond lithium supply. The other moves beyond the liquid electrolyte. They solve different problems, carry different risks, and sit at very different points on the path from paper to product. If you only take one thing from this post, take that.

Below I've sorted the evidence the way I would for a clinical question, adapted for a field that has no meta-analyses in the medical sense. The strongest evidence is commercial-format cells tested under realistic conditions, ideally by someone other than the manufacturer, and ideally in the field. The middle tier is lab prototypes in pouch or multilayer formats with honest reporting of cell-level numbers. The weakest, though often the most exciting, is coin cells, half-cells, and computational screening.

How to read battery evidence

Battery papers routinely report numbers that can't survive scale-up, and it isn't usually dishonesty. A coin cell with a thin electrode, a flooded electrolyte, a large excess of lithium, and a few megapascals of stack pressure can cycle beautifully for a thousand cycles and tell you almost nothing about a 50 Ah pouch cell in a car.

Before trusting a headline figure, I check a few things:

• Is energy density reported per gram of active material, per electrode, or per whole cell? These can differ severalfold.

• What's the areal loading (mAh/cm²)? Thin electrodes flatter rate capability and cycle life.

• How much electrolyte and how much excess lithium or sodium is in the cell? Excess hides consumption that will kill a lean commercial cell.

• For solid-state cells, what stack pressure and temperature were used? Many results depend on conditions no vehicle pack would tolerate.

• What C-rate, and is the cycling at room temperature?

Several journals and groups have pushed reporting checklists for exactly these reasons. Treat any paper that omits these details the way you'd treat a trial that doesn't report its dropout rate.

Sodium-ion: the established picture

Why sodium, and what it costs you

The case for sodium is mostly about materials, not performance. Sodium is vastly more abundant in the Earth's crust than lithium and is widely distributed, which matters more for supply security than raw cost. Sodium doesn't alloy with aluminum at low potentials, so both current collectors can be aluminum instead of using copper on the anode side. That saves cost and weight, and it allows cells to be discharged to zero volts for storage and transport, which is a real safety and logistics advantage.

The physics works against it elsewhere. The sodium ion is larger and heavier than the lithium ion, and sodium's standard electrode potential is roughly 0.3 V less negative. Both cut into energy density. Graphite, the workhorse lithium-ion anode, doesn't store sodium usefully in conventional carbonate electrolytes, so sodium-ion cells mostly rely on hard carbon, a disordered carbon whose storage mechanism (some mix of adsorption, intercalation, and pore filling) is still debated in detail.

The strongest evidence: commercial cells exist

Sodium-ion has crossed the line that matters most. Major cell manufacturers, CATL among them, have announced and moved toward production of commercial sodium-ion cells. CATL's first-generation cell was announced in 2021 at around 160 Wh/kg. Cells in this class have been integrated into small vehicles, two-wheelers, and stationary storage, sometimes in hybrid packs alongside lithium iron phosphate (LFP).

What this tier establishes: commercial sodium-ion cells reach energy densities in the low-to-mid hundreds of Wh/kg, overlapping the lower end of LFP and well below nickel-rich lithium-ion. They show good low-temperature behavior and good rate capability in several chemistries. They can be made on manufacturing lines that are close to existing lithium-ion lines, which is a large part of why they scaled quickly.

What it doesn't yet establish is long-term field reliability across thousands of units and many years. Commercial deployment is recent. The independent, long-horizon degradation data that exist for LFP and NMC simply don't exist yet at comparable scale.

The cathode families

Three cathode families dominate, and each trades something:

• Layered transition-metal oxides give the best energy density but are prone to phase transitions, air and moisture sensitivity, and capacity fade.

• Prussian blue analogues are cheap and have open frameworks that move sodium quickly, but lattice water and vacancies complicate stability and processing.

• Polyanionic compounds (phosphates, fluorophosphates, sulfates) are structurally robust and long-lived but carry lower capacity, and some of the best performers use vanadium, which brings its own cost and supply questions.

The literature here is enormous. A PubMed search on Prussian blue cathodes for sodium-ion or on hard carbon anodes will give you a sense of the volume, and of how much of it is coin-cell work.

The economics are the open question, not the chemistry

My view is that sodium-ion's main uncertainty is no longer technical. It's economic. When lithium carbonate prices spiked in 2022, sodium-ion looked like an obvious hedge. When those prices fell sharply afterward, LFP got cheaper and the cost gap narrowed. Sodium-ion's long-run position depends on whether manufacturing scale and cheap precursors bring its costs well below LFP, or whether it stays a supply-security play that wins in specific niches: cold climates, stationary storage, low-cost vehicles.

Both outcomes are plausible. Anyone telling you confidently which one will happen is forecasting commodity prices, not doing battery science.

Solid-state: the established picture

What the promise actually is

Replace the flammable liquid electrolyte with a solid ion conductor, the argument goes, and two things follow. You might enable a lithium-metal anode, which has far higher specific capacity than graphite and could push cell energy density well beyond today's lithium-ion. And you might improve safety by removing the flammable liquid.

Both claims are conditional, and the conditions are where the research lives.

Electrolyte families and their trade-offs

Polymer systems are the one family with a long commercial record. Bolloré's lithium-metal-polymer batteries ran in car-sharing vehicles for years, heated above room temperature to get adequate conductivity. That's real field evidence, though for a chemistry whose energy density and operating window don't match what most people now picture as "solid-state."

What's firmly established: dendrites don't care that it's solid

The early hope was that a stiff ceramic would mechanically block lithium dendrites. It doesn't, reliably. Lithium penetrates solid electrolytes along grain boundaries, pre-existing cracks, and voids, and it can do so above a critical current density that is often uncomfortably low for fast charging. There's also the opposite problem on discharge: stripping lithium leaves voids at the anode interface, contact is lost, current concentrates at the remaining contact points, and that seeds the next round of penetration. Much of the field's effort is now about interfaces, pressure, and microstructure rather than bulk conductivity. A search on lithium dendrite penetration in solid electrolytes shows how central this has become.

The practical consequence is stack pressure. Many impressive results rely on pressures that are hard to engineer into a pack without adding mass and volume that erode the energy-density advantage. When a paper reports long cycle life, I look for the pressure figure first.

Cathode-side interfaces

Less publicized but just as limiting is the cathode composite. Solid particles touch solid particles imperfectly, active materials expand and contract during cycling, and contacts fail. Sulfide electrolytes decompose against high-voltage cathodes unless the cathode particles are coated. Getting high active-material fractions while keeping ionic and electronic percolation is a materials-processing problem that coin-cell papers often sidestep. The literature on sulfide electrolyte–cathode interfaces is a good place to see how unsolved this still is.

What's emerging, and how much to trust it

Middle-tier evidence: prototype cells

Several companies and large manufacturers have shown multilayer solid-state or anode-free prototype cells, and automakers have announced target windows for limited production in the later 2020s. "Semi-solid" cells, which retain some liquid or gel component, have appeared in a small number of vehicles, mostly in China.

I'd treat this tier the way I'd treat a well-run phase 2 trial: genuinely informative, often sponsor-reported, not yet proof of performance at scale. Watch the terminology closely. "Solid-state" in a press release can mean anything from a fully inorganic cell with a lithium-metal anode to a conventional cell with a modified electrolyte. Those are very different claims.

The safety claim needs more work than it gets

This is the area where I think popular coverage is most ahead of the evidence. Removing the flammable liquid helps, but lithium metal is highly reactive, sulfide electrolytes can produce toxic gas when damaged or exposed to moisture, and thermal-runaway behavior of full solid-state cells at realistic sizes hasn't been characterized nearly as thoroughly as for liquid lithium-ion. Some modeling and abuse-testing work suggests solid-state cells are safer under some failure modes and not inherently safe under others. "Nonflammable" is a property of the electrolyte, not a guarantee for the cell.

Early-stage work: sodium solid-state and beyond

The two lines of research also converge. Solid-state sodium batteries, sodium-metal anodes, and anode-free sodium cells are active areas, along with high-throughput computational screening for new ion conductors in both chemistries. This tier is where the most interesting science happens and where the most claims die. Computational predictions of high conductivity often don't survive synthesis, grain boundaries, and real interfaces. I read this work for ideas, not for timelines.

Where the debates are

The arguments I find most worth following aren't "sodium versus solid-state." They're narrower:

• Is lithium-metal necessary? Some groups argue that solid-state electrolytes with silicon or graphite anodes deliver most of the safety benefit with far less interfacial pain. Others argue that without lithium metal, the energy-density case collapses and the cost premium isn't justified.

• Pressure: engineering or chemistry? One camp treats stack pressure as a pack-design problem. Another argues the interfaces must be made to work at low pressure through interlayers and microstructure control, or the technology won't scale.

• Is LFP the real competitor for both? I think it is. LFP keeps getting cheaper, safer, and denser through cell-to-pack design. Both new chemistries are chasing a moving target, and that rarely gets enough emphasis.

• Hard carbon supply and consistency. Sodium-ion's anode depends on precursors and pyrolysis conditions that strongly affect performance. Whether hard carbon can be made cheaply and consistently at very large scale is an open materials and supply question.

How close is everyday use?

My honest read, by evidence tier:

Sodium-ion is here, in specific applications. It's commercially manufactured and deployed, with lower energy density than mainstream lithium-ion and real advantages in cost structure, cold-weather operation, and shipping safety. Its share depends heavily on lithium prices and manufacturing scale. The missing piece is long-term field data.

Solid-state with a lithium-metal anode is not here for mass-market vehicles. Prototypes are credible and investment is serious, but the hardest problems (interfacial stability, pressure, dendrites at fast charging rates, manufacturing at yield) are still open. Limited, premium, or semi-solid deployments are plausible in the near term. Broad replacement of liquid lithium-ion isn't something I'd bet on soon.

I'd also expect the first everyday encounters most people have with either technology to be unremarkable: a cheaper scooter, a grid storage container, a city car. That's what successful battery chemistries look like. LFP spent years being dismissed as the low-energy option before it took a large share of the market.

What I'd watch next

• Independent testing of commercial sodium-ion cells, especially calendar aging and multi-year field degradation. National labs such as NREL and agencies like the U.S. Department of Energy are natural sources for this kind of non-manufacturer data.

• Solid-state results at low stack pressure in multilayer cells with realistic loadings. When those show up with cycle life and fast-charge data, the timeline changes. Until then, it doesn't.

• Abuse and thermal-runaway studies of full-size solid-state cells, not just electrolyte flammability tests.

• Market analyses that compare against LFP, not against lithium-ion in general. The International Energy Agency publishes battery and critical-minerals analyses that are useful for framing the supply side.

• Reporting standards in the literature itself. The fastest way to improve this field's signal-to-noise ratio would be to stop publishing coin-cell energy densities without cell-level context.

If you're moving into this area from an adjacent field, the volume is the first obstacle. Thousands of papers a year, most of them coin-cell studies, and a handful that matter. I'd build a working corpus early and tag papers by format, loading, and pressure as you read. That's the kind of workflow an AI-native research workspace like PACR is designed for: you search and screen the literature, summarize it, and carry your notes straight into a draft without losing track of which numbers came from which kind of cell.

Whatever tools you use, read the methods section before the abstract. In battery research, the conditions are the result. If you want to start building that corpus now, you can set up a PACR workspace and begin with the searches linked above.