Alkali Metals in Water: The Coulomb Explosion

One shiny lump and a very old demonstration

Cut a fresh piece of sodium and it gleams like mercury for a second before the air fogs it grey. Drop that piece into water and it does something no ordinary metal does: it reacts with the water itself, tearing off hydrogen and leaving a strong base behind. The balanced equation is the same for every member of the family, written here for sodium and in general form:

• 2 Na(s) + 2 H₂O(l) → 2 NaOH(aq) + H₂(g)
• 2 M(s) + 2 H₂O(l) → 2 MOH(aq) + H₂(g)  (M = Li, Na, K, Rb, Cs)

For more than a century, the classroom explanation stopped at "the reaction is exothermic, it produces hydrogen gas, and the hydrogen catches fire." That is all true, and it accounts for the fireball. But it quietly skips the hardest question: why does the reaction accelerate into a bang instead of gently fizzling to a stop? As we will see, the answer is not combustion at all — it is electrostatics, and it happens in the first millisecond, long before any flame.

The redox reaction, split into halves

At heart this is a textbook redox (electron-transfer) reaction. The metal is oxidized and the water is reduced:

• Oxidation (metal loses an electron): Na → Na⁺ + e⁻
• Reduction (water gains it): 2 H₂O + 2 e⁻ → H₂ + 2 OH⁻

Add them and the sodium ions pair with hydroxide to give dissolved NaOH. Track the oxidation states carefully, because this is a common trip-up: the metal goes from 0 to +1, and it is the hydrogen that is reduced, dropping from +1 in water to 0 in H₂. The oxygen never changes — it stays −2 the whole time. No O₂ is involved in making the gas; the bubbles are pure hydrogen.

The reaction is spontaneous by a wide margin. The water half-reaction sits at E° ≈ −0.83 V while the Na⁺/Na couple sits at −2.71 V, giving a cell potential of about +1.9 V. It is also strongly exothermic: forming aqueous NaOH and H₂ releases about −184 kJ for every mole of sodium (equivalently −368 kJ per mole of H₂). That heat is enough to melt sodium on the spot — sodium melts at just 97.8 °C — which is why the metal balls up into a molten, glinting sphere within a fraction of a second. For a moment the reaction also stains the water blue: electrons entering the liquid become solvated (hydrated) electrons, e⁻(aq), the same deep-blue species seen when alkali metals dissolve in liquid ammonia. That fleeting blue is the visible fingerprint of the mechanism to come.

The Coulomb explosion: the first millisecond

The modern picture comes from a 2015 study by Philip Mason, Pavel Jungwirth and colleagues, published in Nature Chemistry. They filmed a droplet of liquid sodium–potassium alloy (NaK, molten at room temperature, so its surface is clean and oxide-free) falling into water using cameras running up to 10,000 frames per second, and they backed the footage with ab initio molecular-dynamics simulations of the metal–water interface.

The simulations showed something startling. Within a few hundred femtoseconds to about a picosecond of contact, the atoms at the metal surface hand their valence electrons over to the surrounding water — far faster than the atoms themselves can move. What is left behind is a skin of bare alkali cations, all carrying the same positive charge, packed together on the surface. Like charges repel, and here the repulsion piles up almost instantly.

The physics that follows is the same instability that governs a charged raindrop or an electrospray needle. A charged liquid surface is held together by surface tension only up to a limit — the Rayleigh limit. Beyond it, electrostatic repulsion overwhelms cohesion and the surface fissions, throwing out fine jets. The freshly positive molten-metal surface blows past that threshold and erupts into metal spikes — sharp filaments of liquid metal launched straight into the water. On the high-speed film these spikes appear in well under a millisecond, far too fast to be debris from a hydrogen explosion.

This is the missing piece. When metal meets water, the heat instantly boils a cushion of steam and the reaction makes hydrogen, wrapping the lump in an insulating gas blanket — a Leidenfrost-like layer that is exactly why a sodium ball skates across the surface instead of reacting all at once. That blanket should throttle the reaction. The Coulomb explosion defeats it: by firing spikes of fresh metal with enormous surface area clean through the gas layer and into liquid water, it exposes new metal faster than the blanket can protect it. Contact area snowballs, heat snowballs, and the reaction runs away into the familiar violent burst.

Why it gets worse down the group

Reactivity climbs steadily down group 1: lithium fizzes politely, sodium melts and may ignite, potassium ignites at once, and rubidium and cesium detonate. Three periodic trends conspire to drive this, and all of them feed the Coulomb-explosion picture rather than fighting it.

• Ionization energy falls. The first ionization energy drops from 520 kJ/mol for lithium to 376 kJ/mol for cesium. The outer electron is held ever more loosely, so the metal sheds electrons into the water faster — building the lethal surface charge sooner.
• Metallic bonding weakens. The same lone valence electron is spread over a larger and larger atom, so cohesion — and melting point — collapse down the group (Li melts at 180 °C, Cs at just 28 °C). A weakly bound, low-surface-tension metal fissions under electrostatic stress far more easily, and it liquefies into a reactive droplet almost instantly.
• Density crosses one. Lithium, sodium and potassium float; rubidium and cesium sink. A sinking metal generates hydrogen beneath the surface and stays fully submerged, which confines the pressure and makes the burst sharper.

There is a famous twist that separates thermodynamics from kinetics. By reduction potential, lithium is the most reactive of all — its E° of −3.04 V is even more negative than cesium's, thanks to the huge hydration energy of the tiny Li⁺ ion. Yet lithium is visibly the gentlest in water. The paradox dissolves once you separate "how far" from "how fast": E° measures the thermodynamic pull, but the rate and the violence are set by kinetics — how quickly electrons leave, how readily the metal melts, and how easily its surface fissions. On those counts lithium is the laggard and cesium the champion.

The fireball: where the hydrogen and the heat come from

Only after the Coulomb explosion has done its work does the classic hydrogen fire arrive. The escaping H₂ mixes with air and ignites, sustained by the reaction's own heat. Hydrogen is unusually easy to light: it burns across an enormous concentration window, roughly 4 % to 75 % by volume in air, and its autoignition temperature is only about 500–570 °C — trivially exceeded at the reaction front. The combustion itself is the familiar

• 2 H₂ + O₂ → 2 H₂O  (≈ −242 kJ per mole of H₂ to steam)

The flame is often tinted by the metal, which is why the demonstration is also a live flame test: sodium burns orange-yellow (589 nm), potassium lilac, lithium crimson, cesium blue-violet. A crucial clarification for the misconception at the top of the page: most of the energy does not come from the hydrogen burning. The bulk of the heat is released by the redox reaction itself — forming the H–H bond of the hydrogen and, above all, the strong hydration of the sodium and hydroxide ions — and that heat exists whether or not the gas ever ignites. The visible fireball is a secondary, if spectacular, encore. Sodium dropped into water under an inert atmosphere still reacts violently; it simply does so without the flame.

Storage, reactors, and doing it safely

Because these metals attack water — and the moisture in air — they cannot be left exposed. Sodium and potassium are stored submerged in mineral oil or kerosene; the far more reactive rubidium and cesium are sealed under argon in glass ampoules. The reactivity is not just a lecture-hall curiosity. Sodium's excellent heat transfer makes liquid sodium a coolant in sodium-cooled fast reactors, where a sodium–water reaction in the steam generator is the defining safety hazard; designers insert an intermediate sodium loop specifically to keep the water and steam a full boundary away from the reactor's sodium. Chemists once dried solvents with sodium wire for the same underlying reason, and quenching that residue with water rather than alcohol has started many a laboratory fire. A metal fire cannot be fought with water or a CO₂ extinguisher — both feed it — so it needs a Class D dry-powder agent that smothers the metal.

The beaker demonstration itself is a controlled experiment: a lentil-sized piece of sodium, done behind a safety shield with eye protection, precisely because the chemistry above guarantees heat, hydrogen and pressure. Understanding the Coulomb-explosion mechanism is not a recipe for making a bigger one — it is the reason the reaction is unforgiving as the piece grows, and the reason it must stay small. The value of the science is knowing why a spoonful of a soft, silvery metal can crack a beaker, so that it is treated with the respect a millisecond-scale runaway deserves.

Reactivity of the group 1 metals with water, from mild lithium to explosive cesium — the trend tracks falling ionization energy and melting point.

Metal	1st ionization energy (kJ/mol)	Melting point (°C)	Behavior in water
Lithium	520	180.5	Steady fizz; floats (0.53 g/cm³), too little heat to melt
Sodium	496	97.8	Melts to a ball, skates on the surface, H₂ may ignite (orange)
Potassium	419	63.5	Ignites at once, lilac flame, sharp pop
Rubidium	403	39.3	Very violent; sinks (1.53 g/cm³) and bursts
Cesium	376	28.4	Explosive — reacts even with ice near −116 °C