In 2007, a paper in Nature reported that photosynthesis might be doing something quantum. The Fenna-Matthews-Olson complex — a light-harvesting protein in green sulfur bacteria — was shown to exhibit oscillating quantum beats in two-dimensional spectroscopy, persisting for more than 660 femtoseconds. The interpretation was striking: excited energy wasn't hopping from chlorophyll molecule to chlorophyll molecule like a classical billiard ball. It was spreading simultaneously across multiple molecules as a quantum superposition, sampling the energy landscape all at once, finding the path to the reaction center by exploring all paths in parallel. Photosynthesis's near-100% quantum efficiency — nearly every absorbed photon resulting in a charge separation event — might be explained by the wave.
The paper went wide in ways scientific papers rarely do. It appeared in TED talks, popular books, a stream of magazine articles about biology discovering what physics had hidden. The idea was compelling: evolution had stumbled onto quantum coherence, turning the wet, warm, noisy interior of a living cell into a quantum computer. It seemed to say something important about the depth of what life had found.
Ten years later, a group at ETH Zürich ran a more careful experiment. The technique, polarization-controlled 2D spectroscopy, can distinguish between two different kinds of oscillating signals that look nearly identical in standard measurements: electronic coherence, where excited electronic states across multiple chromophores are in quantum superposition, and vibrational coherence, where nuclear motions of the molecules are oscillating after being impulsively excited by the laser pulses. The two signals are different physical phenomena. Standard 2D spectroscopy couldn't tell them apart.
The ETH team found that in the FMO complex, electronic coherence dies within 60 femtoseconds. The long-lived oscillations — the hundreds of femtoseconds that had made the 2007 result so exciting — were vibrational. The title of their 2017 PNAS paper left no room for interpretation: "Nature does not rely on long-lived electronic quantum coherence for photosynthetic energy transfer."
This should have settled it. And for several years it more or less did, with theorists arguing that classical vibrational physics could explain the efficiency of photosynthesis without recourse to quantum effects at all. The energy landscape of the FMO complex is simply a downhill funnel — architected by evolution over billions of years to guide excitations to the reaction center reliably. No quantum mystery required.
Then, in October 2025, a computational paper complicated things again. Using a method called DAMPF — dissipation-assisted matrix product factorization, a non-perturbative approach that avoids the approximations all previous theoretical treatments had required — a group found that long-lived excitonic coherences persist in the FMO complex at both liquid nitrogen temperature and room temperature, on picosecond timescales comparable to the energy transfer time. The approximate methods that had supported the revisionist conclusion might themselves have underestimated quantum effects by failing to capture the full dynamics of the system-bath interaction.
Where the field actually stands: genuinely contested. The strongest version of the 2007 claim — that long-lived electronic coherence is functionally essential to photosynthesis — isn't supported. The complete dismissal is now itself under pressure. The truth appears to involve vibronic coupling, a quantum-mechanical mixing of electronic and vibrational degrees of freedom that's a real effect but mechanistically different from either simple interpretation. Whether this coupling helps or whether the efficiency is purely architectural remains unresolved.
The contrast with enzyme hydrogen tunneling is useful here. Tunneling is settled science. When an enzyme transfers a hydrogen atom between substrates, the hydrogen sometimes quantum-mechanically passes through the activation barrier rather than over it. Substituting deuterium (mass 2) for hydrogen drops the tunneling rate dramatically — much more than classical transition-state theory predicts. The H/D kinetic isotope effect in aromatic amine dehydrogenase reaches 55, against a classical maximum of around 7. The anomaly is reproducible, mechanistically understood, and in some cases functionalized: protein residues tens of ångströms from the active site are tuned by evolution to compress the donor-acceptor distance to the range where tunneling becomes efficient. The protein isn't incidentally tunneling. It's been shaped to tunnel.
That's the contrast worth noting. Enzyme tunneling doesn't have the same cultural visibility as photosynthesis quantum coherence — it doesn't lend itself to the same kind of poetic framing — but the evidence for it is considerably harder. It was measured carefully, contested carefully, and the mechanisms were worked out piecemeal over decades. The FMO story caught a larger wave of attention and is still sorting out what exactly happened.
The lesson isn't that quantum biology is overhyped or undersupported. It's that the category "quantum" carries more weight in popular framing than it does in the actual measurements. The Fenna-Matthews-Olson complex doesn't know whether it's performing quantum computation or classical chemistry. It funnels excitation to the reaction center in a few picoseconds. The argument about what to call the mechanism is ours, and it's still running. The bacteria have been doing whatever they do for a billion years regardless.