Elephant Toothpaste: How a Catalyst Turns Hydrogen Peroxide Into a Foam Eruption

What you are actually watching

Strip away the theater and elephant toothpaste is a single reaction running unusually fast: the decomposition of hydrogen peroxide, 2 H₂O₂ → 2 H₂O + O₂. Every molecule of peroxide that breaks apart releases oxygen gas. Without help this happens so slowly you would never notice; with a catalyst, a flask's worth of peroxide dumps its oxygen in seconds. The dish soap added to the flask is not part of the chemistry at all — it is a surfactant that catches the oxygen in a mesh of thin liquid films, turning what would be an invisible fizz into a self-supporting foam that climbs and overflows. Because the reaction is exothermic, the foam is genuinely warm, and in concentrated versions it steams: the wisps you see are water vapor driven off by the reaction's own heat.

Two numbers explain the spectacle. First, stoichiometry: two molecules of peroxide make one molecule of O₂, so a solution that is roughly 10 molar in H₂O₂ carries enough peroxide to release about half its molarity in moles of gas. Second, the gas laws: one mole of O₂ occupies about 24.5 L at room temperature. Multiply those together and a single liter of concentrated peroxide can liberate well over a hundred liters of oxygen, foamed up soap-thin into a volume larger still. The 'eruption' is simply that gas demanding room it does not have inside the flask.

One reaction, two speeds: peroxide's slow secret

Here is the puzzle that makes the demo worth teaching: hydrogen peroxide wants to fall apart. Its decomposition is thermodynamically downhill, with a standard enthalpy of about −98 kJ for every mole of H₂O₂ (equivalently −196 kJ per mole of O₂ released), and the entropy rises too because a liquid is turning partly into gas. Both terms make the Gibbs free energy strongly negative. By the rules of thermodynamics, the reaction should go essentially to completion. Yet a sealed bottle of 3% peroxide sits on a shelf for a year and is only slightly weaker. Why?

Because thermodynamics tells you the destination, not the speed. The decomposition is a disproportionation: oxygen in peroxide sits at the unusual oxidation state of −1, and the reaction simultaneously reduces half of it to −2 (in water) and oxidizes half to 0 (in O₂). Pulling that off requires breaking the weak but real O–O bond and reshuffling atoms through a high-energy arrangement — a barrier of roughly 75 kJ/mol for the uncatalyzed pathway. At room temperature only a vanishing fraction of molecules carry enough energy to clear that barrier at any instant, so the reaction crawls. Pure hydrogen peroxide is therefore metastable: poised at the top of a thermodynamic cliff, held back only by kinetics. Trace iron or copper ions, dust, rough surfaces, heat, and light all quietly catalyze it, which is exactly why peroxide is sold in opaque bottles with stabilizers and vented caps — an uncatalyzed liquid slowly making oxygen that must be allowed to escape.

How a catalyst pulls the trigger

A catalyst does not change where the reaction ends up or how much energy it releases — the enthalpy stays −98 kJ/mol whether you wait a year or a second. What it changes is the path. A catalyst offers an alternative route to the same products through a series of intermediates whose highest energy barrier is lower than the uncatalyzed one. Lower barrier, more molecules over it per second, faster reaction. This is the meaning of the Arrhenius equation, k = A·e^(−Eₐ/RT): the rate depends exponentially on the activation energy, so even a modest cut in Eₐ produces an enormous jump in rate.

The arithmetic is dramatic. Drop Eₐ from about 75 kJ/mol to catalase's ~23 kJ/mol at room temperature and the exponential factor changes by roughly e^((75000−23000)/(8.314×298)) ≈ e^21 ≈ 10⁹. A billion-fold speed-up converts 'years' into 'a fraction of a second' — which is precisely what you see. Crucially, the catalyst is regenerated, not consumed: it is chemically reborn at the end of each cycle and goes on to process molecule after molecule. That is why a pinch of potassium iodide or a spoon of yeast can decompose an entire flask of peroxide, and why the black MnO₂ powder can be filtered out afterward, dried, and used again. The catalyst is a matchmaker that walks away from every wedding intact.

Three ways to light the fuse: iodide, manganese dioxide, and catalase

The same overall reaction can be catalyzed by chemically very different species, and comparing them is the real payoff of the demo.

• Potassium iodide (homogeneous). The iodide ion works through a clean two-step cycle. First, the slow, rate-determining step: H₂O₂ + I⁻ → H₂O + IO⁻ (hypoiodite). Then a fast step regenerates the catalyst: H₂O₂ + IO⁻ → H₂O + O₂ + I⁻. Add the two and the intermediate cancels, leaving the net 2 H₂O₂ → 2 H₂O + O₂ with iodide untouched. Because both peroxide and iodide appear in the slow step, the rate law is rate = k[H₂O₂][I⁻] — first order in each — so doubling the catalyst or the peroxide roughly doubles the burst. Some iodide is oxidized further to I₂ as a side reaction, which is why the KI version's foam is often tinged yellow-brown.
• Manganese dioxide (heterogeneous). Here the peroxide never meets a dissolved catalyst; it adsorbs onto the surface of the solid, where manganese cycles between oxidation states to break the O–O bond, and O₂ desorbs. Because the action is confined to a two-dimensional surface, rate scales with catalyst surface area, and the intact solid can be recovered — a vivid proof that a catalyst is not used up.
• Catalase (enzymatic). Yeast, blood, and liver contain catalase, a heme enzyme built around an iron atom, and it is one of the fastest enzymes ever measured. It uses a ping-pong mechanism: Enz-Fe(III) + H₂O₂ → Compound I + H₂O (an oxidized iron–oxo species), then Compound I + H₂O₂ → Enz-Fe(III) + H₂O + O₂. A single catalase molecule can process on the order of 10⁶–10⁷ H₂O₂ molecules per second, running near the diffusion limit — evolution's answer to the peroxide that its own metabolism produces. This is the catalyst behind the safe, kid-friendly 3% 'yeast' version of the demo.

From bubbles to eruption: the physics of the foam

Chemistry makes the oxygen; physics makes the show. The surfactant in dish soap lowers the surface tension of water and stabilizes thin films, so each escaping O₂ molecule is captured inside a bubble rather than lost to the air. Millions of bubbles pack into a low-density foam whose volume vastly exceeds the liquid that spawned it. Run the numbers for a concentrated demo: a liter of ~30% H₂O₂ is roughly 10 mol/L, yielding about 5 mol of O₂ ≈ 120 L of gas at room temperature — a hundredfold expansion before soap even inflates it. A liter of drugstore 3% peroxide (~0.9 M) yields only about 11 L of gas, which is why it produces a gentle mound instead of a geyser. Concentration is the throttle.

The warmth and steam come from the energy budget. That same liter of 30% peroxide releases on the order of 10 mol × 98 kJ ≈ 1,000 kJ. If none of it escaped, roughly a kilogram of solution would heat by well over 100 °C — so in practice the mixture climbs to 50–80 °C and beyond, foaming and steaming as the heat drives off water vapor. The 3% version releases about a tenth as much and merely feels warm. Everything you observe — how tall the column, how hot the foam, how fast the launch — traces back to the two dials of peroxide concentration and catalyst amount, exactly as the rate law and stoichiometry predict.

Concentration, hazard, and the 'Devil's Toothpaste'

The difference between a birthday-party demo and a laboratory hazard is one number: the peroxide concentration. Household peroxide is 3%; the classic 'monster' demo uses 30–35% laboratory or 'food-grade' grade; the viral 'Devil's Toothpaste' videos use even more concentrated peroxide (well above 30%) to produce a violent, self-heating column. At 30% and above, hydrogen peroxide is corrosive — it whitens and burns skin on contact — and the reaction becomes genuinely energetic. This article explains the science of that eruption precisely so it is understood, not reproduced casually: high-concentration versions are controlled demonstrations done with eye protection, gloves, and containment, and the value here is understanding why the numbers turn dangerous.

The danger is instructive. First, the copious O₂ produced is not itself flammable, but it is a powerful oxidizer: it enriches the local atmosphere so that a glowing splint relights — the textbook test for oxygen — and any nearby fuel burns hotter. Second, the reaction is self-accelerating: it makes heat, heat raises the rate (Arrhenius again), and a large concentrated batch can run away faster than it can shed energy, which is how 'toothpaste' becomes a scalding jet. The very same kinetics that industry harnesses deliberately — concentrated H₂O₂ decomposed over a catalyst is a rocket monopropellant, the basis of hydrogen-peroxide thrusters and, historically, jetpacks and the wartime Walter engines — is what you are watching in miniature. Understanding activation energy and the exponential rate law is exactly what tells you why 3% is a party trick and 90% is propellant.

One reaction, many triggers — how different catalysts speed up the same H₂O₂ decomposition

Catalyst	Type	Activation energy Eₐ	Behavior in the demo
None (pure H₂O₂)	Uncatalyzed	~75 kJ/mol	Decomposes over months to years; the shelf-life problem
Iodide, I⁻ (KI / NaI)	Homogeneous (ionic)	~57 kJ/mol	Two-step hypoiodite cycle; fast burst, often a yellow-brown tint
Manganese dioxide, MnO₂	Heterogeneous (surface)	~50 kJ/mol	Vigorous; the black solid is recovered by filtration afterward
Colloidal platinum	Heterogeneous (metal)	~49 kJ/mol	Very active metal surface; classic kinetics benchmark
Catalase (yeast / liver)	Enzyme (heme iron)	~23 kJ/mol	Fastest of all; kcat up to ~10⁷ s⁻¹; the safe 3% 'yeast' version