Hot Ice: How a Supersaturated Solution Freezes in an Instant

What "hot ice" actually is

Despite the name, hot ice is not ice at all — no water is freezing. It is sodium acetate trihydrate, CH₃COONa·3H₂O, the sodium salt of acetic acid (vinegar) crystallized together with three water molecules per formula unit. The trihydrate has a melting point of about 58 °C. Above that temperature it exists as a clear liquid; below it, the stable form is the white crystalline solid.

The trick of the demonstration is to prepare that liquid and then hold it below its melting point without letting it solidify. A liquid kept below its freezing point (or a solution holding more solute than it should) is metastable: thermodynamically it “wants” to be solid, but it is kinetically stuck, waiting for a trigger. When the trigger arrives, the substance crystallizes almost explosively, and because its melting point is well above room temperature, the newly formed solid is warm — the “hot” in hot ice. Sodium acetate itself is cheap, non-toxic, and food-safe (it is the additive E262), which is why this is a classroom and kitchen-table demonstration rather than a lab-only one.

Building the metastable liquid: supersaturation and supercooling

Sodium acetate is very soluble, and crucially its solubility rises steeply with temperature: roughly 46 g per 100 g of water at 20 °C but well over twice that near boiling. The reason is thermodynamic — dissolving the trihydrate is endothermic (ΔH_solution > 0), so by the van 't Hoff / Le Chatelier logic, heating shifts the equilibrium toward more dissolved salt: d(ln s)/dT = ΔH°/RT² > 0.

To make hot ice you dissolve the salt to saturation in hot water (or simply melt the trihydrate in its own water of crystallization above 58 °C), then let it cool slowly and undisturbed. As the temperature falls, the equilibrium solubility drops below the amount actually dissolved. The excess salt should crystallize out — but with nothing to crystallize onto, it stays in solution. The result is a supersaturated solution holding two to three times its room-temperature equilibrium load. Viewed the other way, the same fluid is a supercooled melt of the trihydrate, sitting some 35–40 °C below its 58 °C melting point. Supersaturation and supercooling are two descriptions of the one metastable state: a system pushed past a phase boundary that has not yet found a way across.

Why nothing happens until you touch it: the nucleation barrier

Metastability survives because starting a crystal from scratch is hard. To grow, the solid must first form a tiny cluster, and creating that cluster costs energy at its surface even as it saves energy in its interior. Classical nucleation theory captures the trade-off for a spherical cluster of radius r:

• A favorable volume term that scales as −r³ (the bulk solid is lower in free energy than the metastable liquid).
• An unfavorable surface term that scales as +r² (the new solid–liquid interface has surface energy γ).

Their sum peaks at a critical radius r* = 2γ/ΔG_v, guarding a free-energy barrier ΔG* ∝ γ³/ΔG_v². Clusters smaller than r* (typically ~1 nm) tend to redissolve; only those that fluctuate past r* keep growing. Spontaneous (homogeneous) nucleation is so improbable here that a clean, undisturbed solution can sit metastable for hours or days.

A seed crystal demolishes the barrier. It presents a ready-made lattice already far larger than r*, so incoming ions simply dock onto an existing surface — heterogeneous nucleation with essentially no barrier. A speck of dust, a scratch on the glass, or the flexed metal disc inside a hand-warmer pack all do the same job, which is why the phrase “touching it with a seed” is literal: contact supplies the missing template.

The crystallization front and the balanced process

Once nucleation begins, a visible solidification front sweeps outward from the seed at the order of centimeters per second. The overall change is a phase transition, not a new-compound reaction — the same substance appears on both sides — and it can be written as ions locking into the hydrate lattice:

Na⁺(aq) + CH₃COO⁻(aq) + 3 H₂O(l)  →  CH₃COONa·3H₂O(s) + heat

Two features make it dramatic. First, the process is self-propagating: each newly formed crystal face is itself a perfect seed for its neighbors, so the front never stalls for want of nuclei. Second, it is autothermal — the heat it releases (next section) keeps the advancing boundary near the optimum growth temperature. Note that the crystal pulls water of crystallization out of the solution and fixes it in the lattice as the trihydrate; the “liquid” genuinely becomes solid rather than merely gelling.

Because solidification happens on contact with a seed, you can pour the supersaturated liquid onto a small crystal and it freezes exactly where it lands, stacking into a self-supporting tower or stalagmite. The liquid solidifies faster than it can spread, so the column grows upward instead of pooling.

Why the solid is hot: latent heat and the thermodynamics

Melting absorbs energy; crystallizing gives it back. The heat released as the trihydrate solidifies is its latent heat of fusion, roughly 264 kJ/kg (about 36 kJ per mole of the 136 g/mol hydrate; literature values span ~226–289 kJ/kg). That energy had been stored when you dissolved or melted the salt with heat; crystallization is simply the bookkeeping running in reverse: ΔH_cryst = −ΔH_fus ≈ −264 kJ/kg.

Thermodynamically the drive is the Gibbs free energy. Below the 58 °C melting point, ΔG = ΔH − TΔS is negative for crystallization: ordering the ions lowers entropy (ΔS < 0), but the large exothermic ΔH dominates at these temperatures, so the solid is the stable phase. Only the nucleation barrier — a kinetic obstacle, not a thermodynamic one — had held the metastable liquid back.

The temperature rise is self-limiting. As latent heat pours into the surroundings, the solid warms — but it cannot climb past its own melting point of ~58 °C, because any hotter and the crystal would simply remelt and reabsorb the heat. So the mass plateaus at roughly 54–58 °C, hot enough to feel distinctly warm but capped by the same phase equilibrium that defines the material.

From demonstration to hand warmers and heat batteries

The demonstration is reusable: gently heat the solid mass above 58 °C (a hot-water bath or microwave) until every crystal redissolves, remove any stray seed crystals from the walls, and let it cool undisturbed. It is metastable again, ready for the next trigger. In practice the cycle degrades only slowly — through water loss from an unsealed container or contamination that seeds premature crystallization.

Sodium acetate trihydrate is a workhorse phase-change material (PCM), prized because it stores a lot of energy at a single, useful temperature:

• Click hand warmers / heat packs. A sealed pouch of supersaturated solution and a small ridged metal disc; flexing the disc pops off micro-crystals that nucleate the pouch, which then holds ~50–54 °C for tens of minutes. Recharge by boiling.
• Thermal energy storage. Because it banks ~264 kJ/kg as latent heat at a fixed 58 °C, it is used to buffer solar-thermal and building-heating systems, releasing stored warmth on demand when nucleated.
• Medical and food warming. Reusable therapeutic hot packs and self-heating containers exploit the same on-demand, controllable heat.

A word on safety: this is a controlled demonstration, and the value here is in understanding why it works. The real hazards are ordinary thermal ones — the solid can reach ~58 °C, warm enough to irritate skin on prolonged contact, and recharging means handling near-boiling liquid. The chemistry itself is benign; sodium acetate is the mild, vinegar-derived salt found in food and buffers.

Common misconceptions

• “It is water freezing.” No — nothing about H₂O→ice is happening. It is a salt hydrate crystallizing; the water present is bound into the trihydrate lattice, not frozen. The name is purely visual.
• “It is a chemical reaction that makes something new.” It is a physical phase change. The identical compound, CH₃COONa·3H₂O, exists before and after; no new molecule forms and no bonds within the acetate are broken. The heat is latent heat of fusion, not a reaction enthalpy.
• “The cold makes it freeze.” Cooling only sets up the supersaturation; the metastable liquid can sit cold indefinitely. What actually triggers solidification is nucleation — a seed, scratch, or metal disc — not a further drop in temperature.
• “Freezing should feel cold.” Freezing (crystallization) is always exothermic; melting is what absorbs heat. Water freezing also releases heat — you just never notice, because it happens at 0 °C. Sodium acetate crystallizes at 58 °C, so its released heat leaves the solid warm.
• “Supersaturation and supercooling are different phenomena.” For hot ice they describe the same metastable fluid — one framed as excess dissolved solute, the other as a melt held below its freezing point.

"Hot ice" versus the ordinary ice it imitates

Feature	"Hot ice" (sodium acetate trihydrate)	Ordinary ice (water)
What crystallizes	CH<sub>3</sub>COONa&middot;3H<sub>2</sub>O salt hydrate	H<sub>2</sub>O molecules
Solidifies at (its melting point)	~58 &deg;C	0 &deg;C
Heat on solidifying	Exothermic, releases ~264 kJ/kg	Exothermic, releases ~334 kJ/kg
Resulting temperature	Warms toward ~54&ndash;58 &deg;C	Stays at 0 &deg;C
Why it stayed liquid	Supersaturated / supercooled, no nucleus	Supercooled, no nucleus
Trigger	Seed crystal or scratch (nucleation)	Seed crystal, dust, or agitation