A barrier that should not be crossed, and a particle that crosses it anyway
Inside your liver, right now, an enzyme called alcohol dehydrogenase is moving hydrogen atoms from ethanol to a coenzyme. The motion is fast. By any classical calculation of how fast it should be — accounting for the height of the energy barrier between reactants and products, the temperature of the cell, and the rate at which thermally-energetic molecules should be able to climb the barrier — the actual rate is much higher than predicted. Sometimes by factors of dozens, sometimes by factors of thousands.
The classical calculation is not wrong about the barrier or the temperature. It is wrong about the assumption that the hydrogen has to go over the barrier. Sometimes it goes through.
This is quantum tunnelling, and it is not an exotic effect imported from physics. It is a routine biological mechanism. The same physics that lights the sun by allowing protons to fuse despite the electrostatic barrier between them is at work in your liver tonight if you have had a drink. The same physics is at work in every cell in your body that respires, in every plant that photosynthesises, and — as the next module will show — in the stacked bases of DNA itself.
This module introduces tunnelling as a quantum-biological mechanism, and shows how proteins do not merely tolerate it but actively promote it through their own vibrational motion. The lesson sets up the destination: charge transport along DNA, where the same principles operate over the densely-stacked aromatic ring systems of the bases.
Tunnelling is too small and too fast to film. The videos here show animations of the mechanism, the experimental signature (the kinetic isotope effect), and how the protein dynamics piece fits together. The aim is to get the picture in your eye before the prose walks you through it.
"Several enzymes use another weird quantum trick called tunnelling to accelerate biochemical reactions. The enzyme promotes a process whereby protons vanish from one position in a biochemical and instantly rematerialise in another." — Johnjoe McFadden, on the work of Judith Klinman (Berkeley) and Nigel Scrutton (Manchester), the experimentalists whose kinetic isotope measurements established hydrogen tunnelling in enzymes.
Read it from substrate to product.
The substrate arrives. A molecule of ethanol (or another alcohol) docks in the active site of alcohol dehydrogenase. The geometry of the enzyme positions the ethanol's hydrogen — specifically the hydrogen on the carbon next to the OH group — within a few Ångströms of a carbon on the NAD⁺ coenzyme. The two carbons are the donor and acceptor; the hydrogen has to move from one to the other.
The barrier. Between the donor and acceptor positions there is an energy barrier — the activation energy. Classically, the hydrogen needs enough thermal energy to climb to the top of the barrier before it can roll down the other side. At body temperature, only a small fraction of molecules have enough thermal energy at any moment. The rate of classical reaction is set by this fraction and by the height of the barrier.
The tunnel. But the hydrogen is a quantum-mechanical particle. Its wavefunction extends through the barrier — exponentially decaying, but not vanishing. There is a non-zero probability that the hydrogen, found near the donor at one moment, is found near the acceptor at the next, even though it never had enough energy to climb. This probability depends sensitively on the width of the barrier and on the mass of the particle. For hydrogen — the lightest atom in chemistry — the wavefunction extends far enough that tunnelling can dominate over the classical climb.
The protein vibrates. The enzyme does not just sit there. It vibrates. Slow protein motions — frequencies in the hundreds of wavenumbers, periods in tens or hundreds of femtoseconds — momentarily bring the donor and acceptor closer together. The barrier between them narrows. The tunnelling probability, which depends exponentially on barrier width, increases by orders of magnitude during these moments. Most tunnelling events happen during these vibrational "windows", not at random.
The product leaves. The hydrogen is now on the NAD⁺, which has become NADH. The ethanol has become acetaldehyde. The enzyme returns to its starting configuration, ready for another substrate. The cycle has turned. A reaction that would have been forbiddingly slow has been accelerated by a billion-fold, partly by the enzyme's geometry (the classical contribution) and partly by tunnelling promoted by vibration (the quantum contribution).
Notice three things about this loop. One: tunnelling is mass-dependent. Lighter particles tunnel more readily; this is why hydrogen tunnels but oxygen practically does not. Replace the hydrogen with deuterium (twice the mass) and the rate drops dramatically — this is the kinetic isotope effect, and it is the standard experimental signature that tunnelling is operating. Two: the protein is not a static catalyst. Its slow vibrations are part of the chemistry. The enzyme is a dynamical machine that prepares quantum events. Three: this physics is not optional. Hydrogen transfer reactions in biology routinely have rates that classical kinetics cannot explain. Tunnelling is the rule, not the exception.
A particle's wavefunction does not stop abruptly at an energy barrier. It penetrates the barrier — decaying exponentially with distance, but with finite amplitude on the other side. The probability of being found beyond the barrier is the square of that amplitude. For barriers narrow enough and particles light enough, the probability is significant. Tunnelling is responsible for radioactive alpha decay, for the operation of every scanning tunnelling microscope, for nuclear fusion in stars, and — as this module establishes — for routine enzyme catalysis. It is one of the foundational consequences of quantum mechanics, and it is doing work in your cells right now.
Where: in the hydrogen transfer step of the enzyme reaction. The hydrogen is light, the barrier is narrow, the tunnelling probability is significant.
Standard textbook reaction kinetics — the Arrhenius equation, the transition-state theory of chemical reactions — is built on a classical picture. In the textbook treatment, the rate is set by how often thermal motion produces a molecule with enough energy to surmount the activation barrier. This treatment is correct for many reactions; it is incomplete for reactions involving the transfer of light particles (hydrogen, protons, electrons) over short distances. In those cases the classical rate is the lower bound; the quantum-mechanical rate, including tunnelling, is higher. The point is not that the classical treatment is wrong — it is that the classical treatment was always an approximation, and in biology the approximation breaks down in routine ways.
Where: in any biological reaction involving hydrogen, proton, or electron transfer over distances of one or two Ångströms. Which is most of them.
The enzyme's role is not to lower the barrier in the static sense imagined by textbook treatments (the "lock and key" image, where the enzyme just holds the substrate in the right orientation). The enzyme also vibrates, with slow modes that periodically bring the donor and acceptor atoms closer together. The tunnelling rate depends exponentially on barrier width, so even small decreases in distance — a fraction of an Ångström — produce large increases in rate. Most tunnelling events occur during these vibrational "windows", not at the average distance. The enzyme is a dynamical regulator of a quantum event: it raises the rate by selectively creating the conditions for tunnelling, then steps back to its average configuration.
Where: in the low-frequency vibrational modes of the protein backbone surrounding the active site. These are the slow, breathing motions of the structure, not the fast bond vibrations.
The same physics that moves a hydrogen atom across a few Ångströms in alcohol dehydrogenase moves an electron across one to two nanometres in an electron transport chain. In mitochondria, the cytochrome chain transports electrons through a sequence of metal centres (iron-sulphur clusters, haem groups, copper sites). Each step is a tunnelling event: an electron jumps from one centre to the next without traversing the protein in any classical sense. The protein scaffold sets the distances and orientations. The chain's efficiency is set by the tunnelling rate at each step, which depends exponentially on the distance — which is why the chain has evolved to keep adjacent centres at 1–2 nanometres apart. Too close: the centres would be electronically coupled, and the chain would not have discrete steps. Too far: tunnelling would be too slow, and the chain would not function.
Where: in the inner mitochondrial membrane, between every adjacent pair of redox centres. Also in the photosynthetic electron transport chain — and, by the same physics, anywhere else where charges hop discretely along a protein.
An electron transport chain is a sequence of discrete tunnelling steps along a protein scaffold. The stacked aromatic bases of DNA are a sequence of strongly-coupled electronic sites along the helical axis. The physics is closely related; the difference is that DNA's bases are more strongly coupled than mitochondrial metal centres, which means charge can move through them not just by discrete hopping but, over short distances, by coherent transport — the wave-like mode of Module Two. This is the bridge to Module Four. Module Four reads the G-quadruplex as a structure that takes this principle and turns up every parameter: tighter stacking, stronger coupling, more symmetric geometry. The cytochrome chain is the warm-up. The quadruplex is the destination.
Where: in the architectural similarity between an electron transport chain and a stack of aromatic rings. Both are quasi-one-dimensional networks for charge.
Why these arrows. The quantum tunnel (A) is the foundational physics; everything downstream depends on it. Once tunnelling is taken seriously, the textbook classical treatment of biological chemistry is exposed as incomplete (A → B). The protein vibrations extend the tunnelling rate beyond what static geometry would predict (A → C). The same physics over longer distances becomes the electron transport chain (A → D, through vibrations C as in marcus theory). The chain in turn is the conceptual model for charge transport in DNA (D → E), which is where Module Four lives.
Serialist: A (the tunnel) → B (the implication for chemistry) → C (the protein's role) → D (the longer-distance case) → E (the bridge forward). The chain works left-to-right.
Holist: Start at the integration — life uses tunnelling routinely — and ask what would have to be true. There must be a quantum mechanism (A); it must matter in chemistry (B); it must work in real proteins (C); it must scale to longer distances (D); it must connect to the rest of biology (E).
What a good answer reproduces: The wavefunction of a particle penetrates an energy barrier with an amplitude that decays exponentially with distance, where the decay rate depends on the square root of the particle's mass and on the barrier height. Lighter particles have wavefunctions that extend further; narrower barriers leak through more. The mass dependence is roughly exponential in the square root of mass (so doubling the mass reduces the rate but not as dramatically as halving the barrier width would). A good answer reaches for the image of a wave decaying through a forbidden region, and recognises why hydrogen is at the boundary where tunnelling is biologically significant: heavier atoms are too heavy, lighter particles (electrons) tunnel easily, but hydrogen sits in the regime where tunnelling adds substantially to the rate without dominating it.
What a good answer reproduces: The textbook picture is correct as far as it goes — enzymes do stabilise the transition state, activation energy is real, and tunnelling is sometimes small. The case that breaks the textbook view is hydrogen transfer over short distances at biological temperatures, where the kinetic isotope effect (the ratio of rates between hydrogen and deuterium) is much larger than the classical maximum of about 7. Soybean lipoxygenase has been measured with kinetic isotope effects above 80; this is unambiguously beyond the classical regime and can only be explained by tunnelling. A good answer names a specific case where the classical picture quantitatively fails, rather than arguing in general terms — and notices that the textbook view is not wrong, it is incomplete.
What a good answer reproduces: Tunnelling rate depends exponentially on barrier width. A static enzyme that held donor and acceptor at the average distance would give the average tunnelling rate. A vibrating enzyme periodically brings them closer; even though it also periodically pushes them apart, the exponential dependence means the rate gain from the close moments overwhelms the rate loss from the far moments. The vibration matters not because it changes the average distance but because it samples short-distance configurations more effectively than a static average. A good answer notices the role of the exponential nonlinearity — and recognises that the protein has been selected, evolutionarily, for vibrations with the right frequency and amplitude to maximise tunnelling.
What a good answer reproduces: Tunnelling rate decays exponentially with distance — roughly by a factor of 10 for every Ångström added. Doubling 1 nm to 2 nm could slow the rate by a factor of 10^10. Halving 1 nm to 0.5 nm would not just speed it up but would couple the centres electronically — they would no longer be discrete sites but a single delocalised system, and the chain would lose its sequential character. The 1–2 nm range is the sweet spot for fast tunnelling between discrete sites. A good answer notices that the chain has been evolutionarily tuned to this range, and that the architectural constraint (centres at 1–2 nm apart) is a direct expression of the underlying physics.
What a good answer reproduces: The standard experiment is the kinetic isotope effect — measure the rate with the natural-abundance substrate (mostly H), then with deuterium (D) and tritium (T) at the transferring position. If the rate ratios fit classical theory, the reaction is classical. If the ratios are much larger than classical theory predicts, or have unusual temperature dependence (decreasing more slowly than Arrhenius at low temperature), tunnelling is contributing significantly. Additional signatures: measuring how the rate depends on substituents that change the barrier width, or comparing computed rates from molecular dynamics simulations with and without tunnelling corrections. A good answer recognises that the question is empirical — there is a specific experimental signature, and the test is whether the data fit it.
What a good answer reproduces: Similar: a one-dimensional array of electronic sites along a structural axis, with charge moving between them by quantum-mechanical mechanisms. Different: DNA's sites are much closer together (3.4 Å vs 1–2 nm), so the electronic coupling between them is much stronger, so the regime of transport is closer to coherent wave-like motion than to discrete hopping. Different: DNA's sites are organic π-systems, not metal centres, so the energetics are different. The direction of the difference: DNA, over short distances, is a better conductor than a cytochrome chain in the sense that charge can move coherently over several bases as a single quantum event. Over longer distances, hopping takes over and DNA conducts more like a cytochrome chain. A good answer notices that DNA combines the quasi-one-dimensional architecture of a chain with the tight stacking of a conjugated system — which is exactly what Module Four will exploit.
What this challenge is for: Pask's meta-conversation. The prediction: Module Four will present a case (the G-quadruplex) where classical chemistry can describe the static structure but cannot account for its regulatory dynamics. Specific evidence (about charge transport, about coherence, about geometric selectivity) will force the same reckoning. The biological structure will turn out to actively use quantum effects, not merely host them. If the pattern holds, the destination is not a surprising new phenomenon but the natural conclusion of the three preceding cases. A learner who can predict this is one who has the framework, not just the examples.
This module ends here, but the entailment continues.
Toward Module Four. The same physics that lets a proton tunnel between two atoms in an enzyme, and an electron jump between two metal centres in a cytochrome, moves charge along the stacked bases of DNA. In a G-quadruplex, the stacking is unusually strong, the geometry is unusually constrained, and the architecture supports both hopping over long distances and coherent transport over short ones. Module Four reads the structure as the destination this series has been heading for.
Toward the broader implication. If routine enzyme catalysis uses tunnelling, and routine respiration uses tunnelling along electron transport chains, then the picture of biochemistry as classical chemistry plus a sprinkling of quantum exotica is wrong. The picture should be reversed: biochemistry is quantum mechanics, much of which can be approximated classically, with the load-bearing exceptions in light-particle transfer and short-distance electron transport. The classical view is the approximation; the quantum view is the foundation.
Toward an open question. Whether enzymes have been evolutionarily tuned for tunnelling specifically — whether the vibrational modes that promote tunnelling are themselves under selection — is an active research area. The case is strong for some enzymes (soybean lipoxygenase, dihydrofolate reductase) and weaker for others. The next decade is likely to settle it. The reader who has reached this point has the framework to follow that literature.
Continue to Module Four