Oscillating chemical reactions

How can a stirred beaker change colour back and forth instead of settling down?

Mix the ingredients of the Belousov-Zhabotinsky reaction in a beaker, stir it steadily, and it turns from red to blue and back, over and over, for many minutes. Everything about that sentence offends an ordinary chemical intuition. Stirring removes any excuse about one part of the liquid differing from another. Nothing is being added or removed. Reactions are supposed to run downhill and stop.

When Boris Belousov reported something like this around 1951, journals declined it as impossible. The referees were not being obtuse: they were applying a real principle. So the question is which principle, exactly, they were applying, and where it fails to bite.

The principle is that a closed system moves toward equilibrium and its free energy falls monotonically on the way. It cannot go back and forth across equilibrium, any more than a ball can roll uphill.

Now be precise about what is oscillating. The colour tracks the oxidation state of a metal catalyst — with ferroin, red when reduced and blue when oxidised. That is one intermediate species among many. The overall reaction, bromate oxidising malonic acid, marches steadily forward the whole time; the free energy really does fall without interruption. What swings is the composition of the intermediate pool on the way down. So the objection was answered not by finding an exception to thermodynamics but by noticing that the quantity being watched is not the quantity thermodynamics constrains.

That removes the prohibition, but it does not produce a rhythm. What does? Ask what would have to be true of the chemistry for the intermediates to swing rather than glide.

First ingredient: something that amplifies itself. In the accepted Field-Koros-Noyes account there is an autocatalytic step — bromous acid reacts with bromate in a way that produces more bromous acid, each round also converting the metal catalyst to its oxidised form. Autocatalysis means the rate rises with the amount already present, so once it starts it does not creep, it detonates. That is the abrupt colour flip.

Second ingredient: something that stops it. If nothing did, the run would end at once and stay ended. Here the stopper is bromide ion, which scavenges bromous acid and so shuts autocatalysis down. While bromide is plentiful the explosive branch cannot start at all; the system quietly consumes bromate through a slower route.

Third and decisive ingredient: the stopper must arrive late. Notice where the bromide comes from. The oxidised catalyst produced by the autocatalytic burst goes on to react with malonic acid and its brominated derivatives, and those reactions release bromide back into the solution — and reduce the catalyst back to its red form. So the burst manufactures its own inhibitor, but only after passing through an intermediate stage. By the time the bromide arrives, the burst has already run to completion. Bromide then rises, suppresses autocatalysis, and is slowly consumed again until it falls past the threshold where the explosive branch can restart.

That is the whole engine: a fast self-amplifying step, a negative feedback that shuts it off, and a delay between them. Remove any one and the rhythm disappears. Without amplification you get a smooth approach; without the negative feedback, one runaway and silence; without the delay, the two balance each other continuously and the system sits at a steady intermediate value.

Does the beaker oscillate forever? No, and that limit confirms the account. In a closed vessel the malonic acid and bromate are eventually spent, the oscillations grow irregular and stop, typically within tens of minutes to a couple of hours depending on the recipe. Feed the reagents in continuously through a stirred flow reactor and the rhythm persists indefinitely. The oscillation is not a property of the molecules; it is a property of a system held away from equilibrium by a supply of free energy.

And the stirring? It is not incidental either. Stirring forces the whole beaker to share one composition, so the colour changes uniformly. Pour the same mixture into an unstirred thin layer and the same chemistry produces travelling bands and rotating spirals instead, because each patch now runs its own cycle and neighbours couple by diffusion.

The most common misreading is that the reaction "goes forwards and backwards". It does not. Individual steps are effectively irreversible under these conditions, and the system never returns to a state it has left in any complete sense — each cycle leaves less bromate and more carbon dioxide behind. Plotted properly, the trajectory is a spiral descending a free-energy slope, not a pendulum swinging about a resting point.

A second clarification concerns how much of this is special chemistry. Very little. The same three ingredients — fast positive feedback, slower negative feedback, and a delay separating them — generate rhythms in systems that share no chemistry at all with a beaker of bromate: the firing of a heart pacemaker cell, the cycling of predator and prey populations, the glycolytic oscillations in yeast, the daily circadian clock built from a protein that represses its own gene after a lag. The Belousov-Zhabotinsky reaction earns its fame partly because it makes an abstract dynamical condition visible on a bench in ten minutes.

It is also worth being honest about the depth of the account. The three-process skeleton is well established and reproduces the behaviour, but the full mechanism involves a large number of elementary steps, and details of the organic chemistry — particularly which bromomalonic acid pathways release bromide, and how — have been revised more than once.

• How the same reaction in an unstirred dish produces spiral waves, and why spirals appear rather than rings.
• What sets an oscillator's period when no component has a timer, and why some biological clocks are far more temperature-stable than chemistry alone would predict.

Every term the collection defines is gathered in the glossary.