Kinetics
Autocatalysis
A reaction that speeds itself up
Autocatalysis is a reaction in which one of the products is itself a catalyst for the same reaction, so the rate accelerates as product accumulates instead of fading away. The simplest form, A + B → 2B, follows the logistic rate law d[B]/dt = k[A][B] and traces a sigmoidal (S-shaped) curve: a slow induction period, an explosive middle as the product catalyst snowballs, and a plateau as the reactant runs out. This positive feedback drives the burst of the permanganate–oxalic acid reaction, the oscillations and travelling waves of the Belousov–Zhabotinsky reaction, prion replication, and the self-replicating networks invoked in the origin of life.
- Defining traitProduct is its own catalyst
- Rate lawd[B]/dt = k[A][B]
- Curve shapeSigmoidal (logistic)
- Max rate at[B] = ½ of total (inflection)
- Classic demoMnO₄⁻ + H₂C₂O₄, Mn²⁺ catalyst
- Feedback typePositive (self-amplifying)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
The idea: a reaction that catalyzes itself
In an ordinary catalyzed reaction you add the catalyst at the start, it speeds things up, and it survives unchanged to the end. In autocatalysis the catalyst is not something you add — it is something the reaction makes. One of the products is itself a catalyst for the very reaction that produces it. So as the reaction proceeds, the amount of catalyst grows, the rate climbs, and the reaction accelerates itself. This is chemistry's purest form of positive feedback.
The minimal cartoon of autocatalysis is
A + B → 2 B
Here B is consumed on the left but produced twice on the right, so each turn of the reaction leaves a net gain of one B. B is both reactant and product catalyst: it is regenerated and multiplied. If you start with a flask of A and only a trace of B, almost nothing happens at first. But every B made goes on to make more B, and the population of catalyst snowballs.
The rate law and the sigmoidal curve
The rate of A + B → 2 B is second order overall, first order in each species:
rate = k[A][B]
What makes this special is that the rate is proportional to the product concentration [B]. Let the total amount of material be fixed, with [A] + [B] = a₀ (every A that disappears becomes a B). Writing x = [B], the rate becomes
dx/dt = k·x·(a₀ − x)
This is the logistic equation — the same equation that governs limited population growth. Its solution is the sigmoidal (S-shaped) curve
x(t) = a₀ / (1 + e−k a₀ (t − t₀))
The curve has three regimes that you can see by eye:
- Induction (lag) period. When x is tiny, the rate k·x·(a₀−x) ≈ k·x·a₀ is small because x itself is small. The reaction creeps along; an observer might think nothing is happening.
- Explosive middle. Both x and (a₀−x) are appreciable, so their product — and the rate — is largest. The rate maxes out exactly at the inflection point, x = a₀/2, where half the reactant has been converted. This is the visible "burst."
- Plateau. As x → a₀ the reactant (a₀−x) is nearly gone, so the rate falls back toward zero and the curve flattens.
The contrast with a normal reaction is stark. A first-order reaction A → products is fastest at t = 0 and decays exponentially. An autocatalytic reaction is slowest at t = 0 and reaches its maximum rate halfway through. Plotting rate against time gives a bell-shaped pulse rather than a monotonic decline.
Ordinary catalysis vs autocatalysis
| Feature | Ordinary catalysis | Autocatalysis |
|---|---|---|
| Source of catalyst | Added externally at the start | Generated as a product |
| Catalyst over time | ≈ constant | Grows from ~0 toward a maximum |
| Where rate is highest | At the beginning (most reactant) | In the middle (inflection point) |
| Concentration–time curve | Exponential decay of reactant | Sigmoidal (logistic) S-curve |
| Early behavior | Immediate, steady | Induction / lag period |
| Feedback character | None (rate just scales) | Positive feedback (self-amplifying) |
| Sensitivity to seeding | Low | High — a trace of product shortens the lag dramatically |
The canonical demonstration: permanganate and oxalic acid
The textbook autocatalytic reaction is the oxidation of oxalic acid by potassium permanganate in warm dilute sulfuric acid:
2 MnO₄⁻ + 5 H₂C₂O₄ + 6 H⁺ → 2 Mn²⁺ + 10 CO₂ + 8 H₂O
Mix the two and at first almost nothing happens — the solution stays deep purple for tens of seconds even at 60 °C. Then, seemingly all at once, the color drains and the flask goes colorless. The reason is that the product Mn²⁺ (and the Mn(III) species it forms en route) catalyzes the reaction. The direct reaction of MnO₄⁻ with oxalate is sluggish; the Mn²⁺/Mn(III) cycle is fast. So the moment enough Mn²⁺ has accumulated, the rate jumps, more Mn²⁺ is made, and the reaction runs away to completion. Add a pinch of MnSO₄ at the start and the induction period vanishes — direct proof that the product is the catalyst. This experiment also explains a common titration warning: in permanganate titrations of oxalate the first few drops decolorize slowly, then progressively faster as Mn²⁺ builds.
Many other real processes are autocatalytic, and several are unwelcome:
- Acid-catalyzed ester hydrolysis. Hydrolysis of an ester releases a carboxylic acid, which supplies the H⁺ that catalyzes further hydrolysis. The degradation of aspirin (acetylsalicylic acid releasing acetic acid) is a textbook case; this is why old aspirin smells of vinegar.
- Tin pest. The transformation of shiny white β-tin to crumbly grey α-tin below 13 °C is autocatalytic: a patch of grey tin nucleates and catalytically converts the surrounding metal.
- Corrosion. Localized corrosion (pitting, crevice corrosion) often acidifies the pit interior; the acid accelerates the corrosion that produces it.
- Polymer and thermal decomposition. PVC dehydrochlorination liberates HCl, which catalyzes more dehydrochlorination — an autocatalytic runaway that thermal stabilizers are designed to suppress.
From feedback to oscillation
Pure autocatalysis runs once and stops. But couple an autocatalytic step (fast positive feedback) to a slower negative feedback — something that consumes the catalyst or its precursor — and the system can oscillate. The concentrations rise as the autocatalysis fires, then fall as the brake engages, then rise again. This is the heart of oscillating reactions, which appear to violate intuition (a beaker that changes color over and over) but are perfectly consistent with thermodynamics: they sit far from equilibrium, burning a fuel toward equilibrium in pulses rather than smoothly.
The famous Belousov–Zhabotinsky (BZ) reaction — malonic acid oxidized by bromate, catalyzed by cerium or ferroin — has at its core the autocatalytic production of bromous acid, HBrO₂: BrO₃⁻ + HBrO₂ + … → 2 HBrO₂ + …. HBrO₂ makes more of itself until bromide ion, regenerated by a separate channel, shuts the autocatalysis off. The cycle repeats every few seconds to minutes, and in an unstirred dish it organizes into spectacular spiral and target travelling waves. The simplest mathematical caricature, the Lotka–Volterra and "Brusselator" models, both require an autocatalytic term to produce limit-cycle oscillations.
Autocatalysis in biology and the origin of life
Self-replication is autocatalysis by another name: a molecule, or a set of molecules, that catalyzes its own formation. Several of biology's most important processes are autocatalytic:
- Enzyme activation cascades. Some zymogens are activated by their own active form — trypsinogen is cleaved to trypsin, and trypsin then activates more trypsinogen. Blood clotting and the complement system use similar self-amplifying loops.
- Prion replication. A misfolded prion protein (PrPSc) templates the misfolding of normal protein into more of itself — a purely conformational autocatalysis with no genetic information, and the mechanism behind diseases like CJD and BSE.
- Template replication. A nucleic acid strand templating its complement, which then templates copies of the original, is the molecular basis of heredity and of PCR's exponential amplification.
At the origin of life this becomes central. Manfred Eigen's hypercycle and Stuart Kauffman's autocatalytic sets argue that a network of mutually catalyzing molecules could grow exponentially, compete for resources, and undergo a primitive natural selection — all before genes or cells existed. Exponential growth is the signature: only an autocatalytic system can outpace its rivals by feeding on its own products. In this view, autocatalysis is not a chemical curiosity but the bridge from dead chemistry to living, self-propagating systems.
Quantifying the lag and the surge
Because the rate constant in d[B]/dt = k[A][B] is multiplied by a₀ in the logistic exponent, the speed of the surge scales with the total concentration: doubling a₀ roughly halves the time to reach the inflection. The induction period also depends on the seed amount of catalyst: starting with x₀ instead of zero shifts the curve earlier by Δt ≈ (1/(k a₀))·ln((a₀ − x₀)/x₀), which is why even a microscopic seed can collapse a long lag. Temperature enters through k in the usual Arrhenius way, so a modest heating can turn a sleepy reaction into a sudden burst — a genuine hazard in runaway exothermic processes, where autocatalysis plus heat release is a classic recipe for a thermal explosion.
Frequently asked questions
What is autocatalysis?
Autocatalysis is a reaction in which one of the products is a catalyst for the same reaction. As product builds up, there is more catalyst, so the rate climbs instead of falling. The textbook form is A + B → 2B, where B both makes more of itself and survives unchanged, giving the logistic rate law d[B]/dt = k[A][B]. Because the rate depends on the product, the reaction starts slowly (an induction period), accelerates sharply, then plateaus as the reactant A is exhausted.
Why does an autocatalytic reaction make an S-shaped curve?
Early on there is almost no catalyst, so the rate is tiny and the product creeps up — the induction (lag) period. Once enough catalyst exists, each new product molecule makes more catalyst, so the rate grows roughly exponentially through the steep middle. Finally the reactant runs low and the rate collapses, flattening the curve. The integrated result of d[x]/dt = k x (x_max − x) is the logistic (sigmoidal) function — the same S-shape as population growth, with the inflection point at x = x_max/2.
What is a real example of autocatalysis?
The permanganate–oxalic acid reaction, 2 MnO₄⁻ + 5 H₂C₂O₄ + 6 H⁺ → 2 Mn²⁺ + 10 CO₂ + 8 H₂O, is the classic demonstration. The purple permanganate is slow to react at first, but the Mn²⁺ it produces catalyzes the reaction, so the color suddenly drains in a burst. Other examples include acid-catalyzed ester hydrolysis (which releases more acid), tin pest, the hydrolysis of aspirin, and the corrosion of metals where the product acidifies the local environment.
How is autocatalysis different from a normal catalyst?
A normal catalyst is added at the start and stays roughly constant, so the rate is highest at the beginning and falls as reactant is used up. In autocatalysis the catalyst is generated by the reaction itself, so it starts near zero and grows. That flips the kinetics: the maximum rate is in the middle, not at the start, and the curve is sigmoidal rather than a simple decay. Autocatalysis is a form of positive feedback; ordinary catalysis is not.
How does autocatalysis cause chemical oscillations?
An autocatalytic step provides the positive feedback (explosive growth of an intermediate), and a delayed negative feedback or a second consuming reaction resets it. Coupled together, the concentrations rise and fall in cycles. The Belousov–Zhabotinsky reaction is the famous case: the autocatalytic production of HBrO₂ drives a burst, bromide ion shuts it down, and the system oscillates in color (and forms travelling waves) for many minutes far from equilibrium.
Is autocatalysis important for the origin of life?
Yes. Self-replication — the defining feature of life — is autocatalysis: a molecule (or a set of molecules) that catalyzes its own formation. Manfred Eigen's hypercycle and Stuart Kauffman's autocatalytic sets propose that networks of mutually catalyzing molecules could grow exponentially and undergo selection before genes existed. Template replication of nucleic acids and the spread of prions are both real autocatalytic processes.