Sodium-Ion Batteries

June 21, 2026 21 min read Infrastructure of Civilization

sodium-ionbatteriesenergy-storagegrid-storagelfpelectrochemistryevmaterials

Contents

Sodium-ion batteries are the chemistry everyone reaches for when lithium gets expensive and ignores the rest of the time. That is the whole story of the technology’s commercial life so far. When lithium carbonate spiked to roughly $80,000 per tonne at the end of 2022, sodium-ion went from a lab curiosity to a CATL product launch and a dozen Chinese pilot plants inside eighteen months. When lithium then collapsed back below $15,000 per tonne through 2024 and into 2025, every sodium-ion business plan that assumed a permanent lithium premium quietly got rewritten, and at least one high-profile Western program (Northvolt’s) was overtaken by the company’s broader collapse before it shipped anything at scale. The chemistry did not change. The arithmetic did.

This makes sodium-ion an unusually honest case study, because the technology has no headline advantage that survives contact with a spreadsheet. It is not denser than lithium-ion; it is meaningfully less dense. It is not a dramatic safety breakthrough the way solid-state is marketed to be, though it is genuinely better behaved. Its entire pitch is cost and materials abundance, and that pitch is contingent on commodity prices the manufacturer does not control. So the interesting question is not “is sodium-ion good,” which is unanswerable, but “where does the cost-and-abundance argument beat the density penalty,” which has specific, defensible answers: stationary grid storage, cold-climate operation, and low-cost short-range mobility, primarily in China where the supply chain actually exists.

This post walks the electrochemistry — why a sodium ion behaves differently from a lithium ion, why graphite anodes do not work and hard carbon is used instead, and the three cathode families competing for the cathode slot. Then it lays out the genuine advantages (cheap abundant sodium, no cobalt or copper, aluminum current collectors on both sides, shipment at zero volts, cold-weather performance) against the honest disadvantages (energy density, volumetric density, cycle-life maturity). Then the real 2025-2026 deployments, and finally the cost-trajectory honesty that decides everything.

Why sodium is different from lithium

Sodium sits directly below lithium in group 1 of the periodic table, so chemically it is the obvious substitute: both are alkali metals, both give up a single valence electron easily to form a +1 ion, both shuttle that ion between a cathode and an anode in an intercalation cell. The mechanism of a sodium-ion cell is identical in concept to a lithium-ion cell. Charge drives Na+ out of the cathode, through a sodium-salt electrolyte, and into the anode; discharge reverses it. If you understand lithium-ion, you understand the shape of sodium-ion immediately, and the broader chemistry-selection logic in the battery chemistry compared deep dive transfers directly.

The differences are quantitative and they all flow from two physical facts. First, the sodium ion is bigger. The ionic radius of Na+ is about 1.02 angstroms versus 0.76 angstroms for Li+ — roughly a third larger in radius, which is a substantially larger volume. A larger ion is harder to fit into a host crystal lattice, diffuses more slowly through it, and strains the structure more on each insertion and removal. Second, sodium is heavier (23 g/mol versus 7 g/mol for lithium) and its standard reduction potential is less negative (-2.71 V versus -3.04 V for lithium versus the standard hydrogen electrode). The voltage gap means a sodium-ion cell of otherwise identical chemistry runs at a slightly lower cell voltage, typically 0.3 to 0.4 V lower, and the heavier ion means more mass is moved for the same charge. Both effects subtract from gravimetric energy density before you have made a single engineering choice.

These are not defects to be engineered away. They are the fixed background against which every sodium-ion advantage and disadvantage plays out. The larger, slower ion is why the anode has to change. The lower voltage and heavier ion are why the energy density ceiling is lower. And the fact that sodium is the sixth most abundant element in the Earth’s crust, extractable from ordinary soda ash and even seawater, is why anyone bothers despite those penalties.

The anode: why graphite fails and hard carbon wins

In a lithium-ion cell the anode is graphite, and graphite is close to ideal. Lithium intercalates between graphene sheets to form LiC6, storing one lithium per six carbons at 372 mAh/g, with a flat, low potential close to lithium metal. You would naturally try the same thing with sodium. It does not work.

Sodium does not form a stable graphite intercalation compound under normal conditions. The thermodynamics are unfavorable — the binding of sodium between graphene layers is too weak relative to the energy cost of expanding the lattice — so you cannot make a useful NaC6 the way you make LiC6. Sodium intercalates into graphite to a useful degree only with specific co-intercalating solvents (ether-based electrolytes that drag solvent molecules in with the ion), which is an active research path but not what commercial cells do. The practical answer is to abandon the ordered graphite structure entirely and use hard carbon.

Hard carbon is a disordered, non-graphitizing carbon — pyrolyzed organic precursors (biomass, sugars, phenolic resins, coal-derived pitch) that, unlike graphite, never form long-range ordered graphene stacking even at high temperature. Its structure is a tangle of short, curved graphene fragments enclosing nanoscale voids. Sodium stores in two distinct ways here: it inserts into the small spaces between the disordered graphene fragments at higher potential (the sloping part of the voltage curve), and it then fills the nanopores at very low potential close to sodium plating (the flat plateau near 0.1 V). That plateau is where most of the capacity lives and where most of the energy density comes from, because the flat low-voltage plateau is what gives the cell a high average voltage. Good hard carbons deliver 300 to 350 mAh/g, comparable to graphite’s 372 mAh/g, which is one reason the anode-side penalty for sodium is smaller than you would guess.

Hard carbon brings real engineering problems. The disordered structure with all its internal surface area gives a large irreversible first-cycle capacity loss — sodium consumed forming the solid-electrolyte interphase that never comes back — so first-cycle Coulombic efficiency is often 80 to 90 percent versus 90-plus for good graphite, which wastes cathode sodium and costs energy density. The low-voltage plateau sits dangerously close to sodium-metal plating, so charging tolerances and BMS behavior matter. And hard carbon performance is exquisitely sensitive to the precursor and the pyrolysis schedule, so cell makers guard their hard-carbon recipes the way lithium makers guard cathode coatings. None of this is fatal. All of it is why sodium-ion took as long as it did to reach production-quality cells.

The three cathode families

Unlike the anode, where hard carbon has effectively won, the cathode is still a three-way contest. Each family trades energy density against cost, cycle life, and air stability differently, and the major cell makers have split across them.

Cathode family	Representative chemistry	Approx. capacity	Strengths	Weaknesses
Layered transition-metal oxides	NaNi/Mn/Fe/Cu/Ti oxides (O3 / P2 types)	120-160 mAh/g	Highest energy density, drop-in fit to existing electrode coating lines	Air- and moisture-sensitive, complex multi-metal recipes, can use some nickel/copper
Prussian-blue analogues	Na2Fe[Fe(CN)6] and Mn/Fe variants	120-160 mAh/g	Cheapest materials, open framework tolerates large ions, good rate	Low tap density (poor volumetric), residual water and cyanide-handling concerns, manufacturing maturity
Polyanionic	Na3V2(PO4)3, Na vanadium fluorophosphates, NaFePO4F	100-130 mAh/g	Best cycle life and thermal/structural stability, flat voltage	Lower capacity, vanadium versions are expensive, lower intrinsic conductivity needs carbon coating

Layered oxides are the structural analogue of NMC and LFP cathodes — transition-metal-oxide layers with sodium sitting between them — and they give the highest energy density, which is why CATL and most companies chasing automotive-adjacent applications use them. They come in two structural types distinguished by how the sodium sits (O3, where sodium is octahedrally coordinated and the cell starts fully sodiated; P2, prismatic coordination with better rate and stability but a sodium deficit that has to be made up). The catch is air sensitivity: many sodium layered oxides react with atmospheric moisture and CO2, degrading on the shelf, so electrode handling has to stay dry. The recipes are also genuinely complex, deliberately blending four or five transition metals (iron, manganese, copper, nickel, titanium) to stabilize the structure and tune voltage, which is a formulation art rather than a single clean compound.

Prussian-blue analogues are the cost story taken to its limit. These are iron-cyanide framework compounds — the same family as the pigment Prussian blue — with an open cubic structure full of large interstitial sites that the bulky sodium ion slots into comfortably. The raw materials are essentially iron and sodium, about as cheap as cathode materials get. The problems are practical: the open framework has low density, so volumetric energy density suffers; the compounds trap water that has to be driven out carefully; and handling cyanide-bearing precursors at scale is a process-engineering and safety burden. Natron Energy built its product line on Prussian-blue chemistry for high-power, ultra-long-cycle stationary use before winding down in 2024, which says something about both the appeal and the difficulty.

Polyanionic cathodes — phosphates, fluorophosphates, sulfates — are the sodium analogues of LFP, and they inherit LFP’s virtues. The strong covalent polyanion framework is structurally and thermally rock-solid, giving the best cycle life and the best safety in the sodium family, with flat, predictable voltage. The cost is lower capacity and, in the highest-performing vanadium-based versions (Na3V2(PO4)3), an expensive and somewhat toxic transition metal that undercuts the cheap-materials thesis. Iron-based polyanionic cathodes avoid the vanadium but give up energy density. This family is the natural choice for grid storage that has to last twenty years, which is exactly where some Chinese deployments have landed.

How the cell actually works

The shuttle is the same as lithium-ion, with sodium in the carrier role. The one structurally interesting difference is the current collectors.

              CHARGE  (Na+ leaves cathode, enters hard carbon)
   e- ----------------- external circuit / charger -----------------> e-
   |                                                                  |
 [ALUMINUM]                                                      [ALUMINUM]
 cathode                                                          anode
 collector                                                        collector
   |                                                                  |
+------------+        Na+ -->  Na+ -->  Na+                  +---------------+
|  Na layered |     ============================            |  hard carbon  |
|  oxide /    |       sodium-salt electrolyte               |  (disordered  |
|  PBA / poly |       (NaPF6 in carbonate)                  |   graphene +  |
|  cathode    |     ============================            |   nanopores)  |
+------------+        <-- Na+  <-- Na+  <-- Na+              +---------------+
                            DISCHARGE
   <---------------- e- through the load <----------------

  KEY DIFFERENCE vs Li-ion: sodium does NOT alloy with aluminum, so the
  anode collector can be ALUMINUM too -- not the copper a Li-ion cell needs.
  Bonus: a Na-ion cell can be fully discharged to 0 V without dissolving a
  copper collector, so it ships SAFE and inert.

The aluminum-on-both-sides detail is not cosmetic. In a lithium-ion cell the anode current collector has to be copper, because lithium alloys with aluminum at the anode’s low potential and would destroy an aluminum foil. Copper is heavier, more expensive, and a meaningful fraction of cell cost and weight. Sodium does not alloy with aluminum, so a sodium-ion cell uses thin aluminum foil on both electrodes. That removes copper from the bill of materials entirely and shaves cell mass. It also unlocks the zero-volt shipping trick: because there is no copper collector to dissolve, a sodium-ion cell can be discharged all the way to 0 V, where it is electrochemically inert and safe to ship and store, then brought back up with no damage. A lithium-ion cell held at 0 V dissolves its copper collector and is permanently ruined, which is why lithium cells ship at partial charge under hazardous-goods rules. This is a genuine, structural sodium-ion advantage that has nothing to do with marketing.

The genuine advantages

Strip out the hype and a defensible list remains.

Materials abundance and cost floor. Sodium is the sixth most abundant element in the crust and is recovered from soda ash (sodium carbonate), which trades around $300 per tonne versus lithium carbonate’s wildly variable $15,000-to-$80,000-per-tonne range. There is no geographic chokepoint the way lithium has Australia, Chile, and China, and no cobalt to source from the DRC. Combined with the elimination of copper and (in iron-based cathodes) any nickel or cobalt, the raw-material cost floor of a sodium-ion cell is structurally below LFP. The word “floor” is doing work here — see the cost section.

Cold-temperature performance. Sodium-ion cells retain a markedly higher fraction of their capacity at low temperature than lithium-ion. Manufacturers report retaining 85-90 percent of capacity at -20 C where LFP can sag to 60-70 percent, and operation down to -30 or -40 C. The mechanism is favorable sodium-ion desolvation kinetics and electrolyte behavior at low temperature; the practical consequence is that a sodium-ion pack in a cold climate needs less heating energy and delivers more usable capacity in winter, which directly attacks one of LFP’s real weaknesses.

Safety and thermal behavior. Sodium-ion cells are not non-flammable, but they tolerate abuse well, have a less aggressive thermal-runaway signature than high-nickel lithium chemistries, and the zero-volt-safe property makes the entire logistics chain safer. Polyanionic and Prussian-blue cathodes in particular do not release oxygen readily, similar to the structural argument that makes LFP safe.

Fast charge and power. The open frameworks (especially Prussian-blue and some layered oxides) and favorable kinetics give good rate capability, so high C-rate charge and discharge are achievable without the lithium-plating risk that constrains cold-weather lithium charging.

The honest disadvantages

The reason sodium-ion is not simply replacing LFP everywhere is a short, blunt list.

Gravimetric energy density. This is the headline penalty. Commercial sodium-ion cells today land roughly in the 100-160 Wh/kg range, with the best layered-oxide cells now reaching into the 160s and second-generation products claiming higher. LFP is roughly 150-200 Wh/kg at the cell level and NMC well above that, into the 250-300 Wh/kg range for high-nickel cells. For anything where mass is the binding constraint, that gap is decisive and it does not close with a process tweak — it is rooted in the heavier ion and lower cell voltage described above.

Volumetric energy density. Often worse than the gravimetric gap, especially for Prussian-blue cathodes with their low tap density. Volumetric density (Wh/L) matters as much as gravimetric in space-constrained applications, and sodium-ion’s open, lower-density cathode structures lose here. A grid container or a floor-standing home battery has room to spare, so this penalty is survivable in stationary use and painful in vehicles.

Cycle-life and manufacturing maturity. This is the moving target. Polyanionic sodium cells can demonstrate long cycle life on paper, and CATL and others now quote multi-thousand-cycle figures, but the field has roughly fifteen fewer years of mass-production learning than lithium-ion. Real-world cycle life, calendar life, and pack-level reliability data across many gigawatt-hours and several years simply does not exist yet to the depth that LFP’s does. The chemistry can be good; the proof at fleet scale is still being accumulated, and that uncertainty is correctly priced into buyers’ caution.

Metric	Sodium-ion (2025-26)	LFP	NMC (high-nickel)
Gravimetric density (cell, Wh/kg)	100-160	150-200	250-300
Volumetric density (cell, Wh/L)	250-350	350-450	600-700
Cycle life (80% DoD)	2,000-5,000 (claimed; maturing)	3,000-6,000+	1,000-2,000
Cell cost trajectory ($/kWh)	~$45-75, falling with scale	~$50-70 (2025)	~$80-110
Cold performance (-20 C cap. retention)	85-90%	60-70%	70-80%
Anode collector	aluminum	copper	copper
Cobalt / nickel / copper required	none required	none (Cu collector only)	yes
Ship at 0 V	yes (safe/inert)	no	no

The cost row deserves a flag: the sodium-ion numbers are the projected floor at scale, not necessarily what you pay today. In 2025, with lithium cheap and sodium-ion volumes still small, real sodium-ion cell prices were not reliably below LFP. The chemistry’s cost advantage is structural but it is unlocked by scale, and scale is exactly what it has not yet reached outside China.

The real deployments, 2025-2026

The technology is genuinely shipping, and overwhelmingly from China, where the cell supply chain, the hard-carbon supply, and the buyers all sit.

CATL is the bellwether. It launched its first-generation sodium-ion cell in 2021 at around 160 Wh/kg, then in 2025 announced its Naxtra branded sodium-ion line, including a passenger-EV cell targeting roughly 175 Wh/kg and a heavy-truck and grid-oriented product, with mass production scheduled around the end of 2025. CATL’s framing is telling: it positions sodium-ion not as an LFP replacement but as a complement for cold climates, entry-level range, and storage, and it has pushed hybrid packs that combine sodium-ion and lithium cells in one pack (its “AB” architecture) to blend sodium’s cold-weather and cost behavior with lithium’s density.

BYD and Huaihai announced a joint venture in 2024 to build what was billed as one of the largest sodium-ion plants for micro-mobility and small EVs, targeting the enormous Chinese two-wheeler and low-speed-vehicle market — a segment where range is short, cost is everything, and cold-weather start matters. This is sodium-ion’s most natural mobility beachhead: not the long-range car, but the electric scooter and the city runabout.

HiNa Battery, the spinout from the Chinese Academy of Sciences’ Institute of Physics, has been the academic-to-commercial bridge, shipping layered-oxide sodium-ion cells and supplying early grid-storage demonstrations, including multi-megawatt-hour stationary installations in China that served as proof-of-concept for sodium-ion grid storage. A widely cited early milestone was a sodium-ion grid-storage station on the order of single-digit MWh commissioned to validate the chemistry at utility scale, with larger projects following.

Northvolt is the cautionary Western counterexample. In late 2023 it announced a sodium-ion cell (Prussian-blue-based, developed with Altris) at around 160 Wh/kg aimed explicitly at grid storage and emerging markets, a credible technical result. But Northvolt’s broader financial collapse through 2024 into bankruptcy overtook the program before it reached meaningful volume, which is itself a lesson: a good cell does not survive a company that cannot finance its gigafactories.

The pattern across all of this is consistent. The wins are stationary grid storage (HiNa, polyanionic and layered-oxide cells in Chinese utility projects), entry-level and two-wheeler mobility (BYD/Huaihai), and cold-climate or cost-sensitive niches (CATL’s complement positioning). Nobody serious is claiming sodium-ion for the long-range premium EV. That is not where it wins.

Where it wins and where LFP still wins

Stationary grid and home storage. This is the strongest case. A grid battery does not care about weight or volume within reason — a few extra containers on a concrete pad is a rounding error against the land and interconnection costs. What it cares about is dollars per kilowatt-hour, cycle life over a twenty-year financed asset, safety, and supply-chain security. Sodium-ion’s cheap, abundant, conflict-free materials and good cycle life (especially polyanionic) line up exactly with that priority list. If sodium-ion ever beats LFP decisively anywhere, it is here.

Cold climates. The -20 C capacity advantage is real and structural, and it compounds with the storage case: a stationary battery in a cold region delivers more winter capacity and burns less energy heating itself.

Low-cost short-range mobility. Electric two-wheelers, three-wheelers, low-speed urban vehicles, and entry-level city cars where the daily range requirement is modest and the price is the entire purchase decision. The density penalty is affordable when you only need 100-200 km, and the cost and cold-weather behavior are direct wins.

Where LFP still wins: range-sensitive EVs. Any vehicle where customers buy range, the gravimetric and volumetric density gap is simply too large. A sodium-ion pack heavy and bulky enough to match an LFP car’s range erases the cost advantage and then some. This is structural, not a maturity problem, and it will not change. For the density frontier above LFP — and the marketing around it — the solid-state batteries writeup covers what is actually shipping versus announced.

Cost-trajectory honesty: the lithium-price crossover

Everything about sodium-ion’s commercial prospects reduces to one chart: the cell cost of sodium-ion versus LFP as a function of the lithium price. Sodium-ion’s bill of materials is genuinely cheaper at the floor, but two things determine whether that floor is reached in practice — manufacturing scale, and how much lithium costs that quarter.

  Cell cost
  ($/kWh)
    |
110 |  *  NMC (lithium + nickel exposed)
    |   *.
 90 |     *.            LFP cost rises with lithium price --->
    |       *. . . . . . . . . . . . . . . . . . . . . . . .*  LFP
 70 |          *  . . . . . . . . . . . . . . . . . . .* '
    |             * . . . . . . . . . . . . . .* '          <-- CROSSOVER
 50 |================*=========*=========* '================  Na-ion (flat-ish:
    |                  Na-ion barely moves with Li price      almost no Li in BOM)
 30 |
    +----+---------+---------+---------+---------+---------+--> Li2CO3 price
        $10k      $20k      $40k      $60k      $80k     ($/tonne)
                                |
                  2024-25 regime: cheap lithium,
                  LFP sits BELOW or at Na-ion --> Na-ion struggles

                  2022-23 spike: lithium ~$80k,
                  LFP well ABOVE Na-ion --> Na-ion looks unbeatable

In late 2022 and early 2023, lithium carbonate spiked toward $80,000 per tonne. At that price LFP’s cathode cost ballooned and sodium-ion looked like a structural winner — which is precisely when the announcements flooded in and CATL launched product. Then through 2024 and 2025 lithium collapsed back below $15,000 per tonne as new supply came online and EV-demand growth normalized. At cheap lithium, the lithium content of an LFP cell is a small fraction of cell cost, LFP’s price floor dropped to the $50-70/kWh range, and sodium-ion’s structural materials advantage shrank to something that scale-disadvantaged sodium-ion factories could not actually realize. BloombergNEF’s battery-price surveys through this period show the volume-weighted lithium-ion pack price continuing to fall, which is the headwind sodium-ion has to beat.

The honest conclusion is that sodium-ion is not unconditionally cheaper than LFP. It is cheaper at the materials floor, and it becomes cheaper in practice when (a) it reaches LFP-comparable manufacturing scale, and (b) lithium prices are elevated. Both conditions are partly outside any cell maker’s control. This is why sodium-ion’s commercial story has been so volatile: the same chemistry was a sure thing in 2023 and a question mark in 2025, and the only thing that changed was a commodity price. The supply-chain-security argument — sodium has no Chile, no DRC, no single-country chokepoint — is the part of the cost case that does not depend on the spot price, and it is why governments and grid operators keep funding sodium-ion even when the spreadsheet is marginal.

Verdict

Sodium-ion is a real, shipping, well-understood battery chemistry with a narrow but defensible set of applications, and it is not a lithium-ion replacement. The electrochemistry is sound: a larger, heavier, slower ion shuttling between a hard-carbon anode and one of three cathode families, at a slightly lower voltage than lithium-ion, with a structurally lower energy-density ceiling. That ceiling is the permanent fact everything else negotiates around.

Where the chemistry wins, it wins on its own terms — not by beating lithium-ion at density, but by being cheaper-at-the-floor, materials-abundant, cobalt-and-copper-free, cold-tolerant, and safe to ship at zero volts. Those advantages map cleanly onto stationary grid storage, cold-climate operation, and low-cost short-range mobility, which is exactly where the genuine 2025-2026 deployments have landed: CATL’s Naxtra complement strategy and AB hybrid packs, BYD and Huaihai’s two-wheeler plant, HiNa’s grid demonstrations. Where customers buy range, LFP still wins and will keep winning, because the density gap is structural.

The one number that decides sodium-ion’s near-term commercial fate is not a property of the cell at all. It is the price of lithium carbonate. At elevated lithium prices and LFP-comparable manufacturing scale, sodium-ion is a structural cost winner. At the cheap-lithium regime of 2024-2025, with sodium-ion still subscale, the advantage compressed to the point where it competed on supply-chain security and cold-weather performance rather than on raw price. The chemistry is patient; it will sit there being modestly cheaper at the materials floor and waiting for the next lithium spike, the next scale-up, or the next grid operator who values not depending on Chile. For grid storage and the Chinese low-cost mobility market, that wait is already over. For everything else, sodium-ion is a hedge that becomes a winner only when the arithmetic says so.