On a bench inside the James H. Clark Center, home to Stanford Bio-X, a research team spent part of 2024 pointing a solar simulator lamp at a device the size of a postage stamp. The device was not silicon. It was a thin conductive scaffold seeded with photosynthetic material, wired to a picoammeter that logged current in real time. The number that came back, according to the team’s account of the work, was roughly three times higher than the energy conversion rate of the leaf sitting in a planter across the room.
That comparison sounds like it should be hard to achieve. It is not. Plants are, by the standards of energy engineering, remarkably bad at turning sunlight into usable energy. The Stanford result is less a triumph over evolution’s cleverness than a demonstration of how much of a leaf’s captured light gets thrown away before it becomes anything a human could use. Understanding why requires understanding exactly where that light goes after a chloroplast absorbs it, and why a wire stuck directly into the photosynthetic machinery can skip most of those losses entirely.
Photosynthesis Was Never Optimized to Be Efficient
Natural photosynthesis carries a theoretical ceiling of about 4.6 percent for C3 plants, the category that includes wheat, rice, and most trees, and closer to 6 percent for C4 plants like sugarcane and corn. Those numbers describe the fraction of total incoming solar energy that could, in principle, end up stored as chemical energy in a plant’s tissue. In practice, field efficiency for most crops runs closer to 1 percent, sometimes less. A cornfield in July is capturing a sliver of the energy falling on it.
The gap between the theoretical ceiling and the real number is where the story gets interesting. Chlorophyll only absorbs certain wavelengths of visible light efficiently, so a chunk of the solar spectrum is invisible to the plant from the start. Photorespiration, a wasteful side reaction that happens when the enzyme RuBisCO grabs oxygen instead of carbon dioxide, burns off more energy, especially in hot, dry conditions. The Calvin cycle itself, the set of biochemical reactions that fixes carbon into sugar, has its own conversion losses built in. Then the plant spends a meaningful share of whatever energy survives all of that on respiration, root maintenance, and structural growth that never shows up as stored solar energy at all.
Beating plant photosynthesis was never the hard problem. Plants set an extremely low bar.
None of this is a flaw in evolution’s design. Photosynthesis was never selected for energy conversion efficiency. It was selected for reproductive success, drought tolerance, and competition for light and nutrients over millions of years, in environments where surviving a dry season mattered more than maximizing joules per square meter. A plant that photosynthesized at 20 percent efficiency but couldn’t survive a cloudy week would have lost the evolutionary race to one that photosynthesized at 1 percent and reproduced reliably. Efficiency, in other words, was never the metric nature was optimizing for.
What a Bio-Solar Cell Actually Skips
A biohybrid solar cell like the one described in the Stanford work takes a different approach to the same starting material. Instead of letting a photon travel through the full biological chain, from light absorption through the electron transport chain, into the Calvin cycle, into sugar, into biomass, the device intercepts electrons much earlier, often straight out of the photosynthetic electron transport chain itself, and routes them directly to an electrode.
This is the same underlying idea behind biophotovoltaic research done over the past two decades at institutions including Vanderbilt, where chemists have built solar cells using isolated Photosystem I protein complexes bonded to conductive surfaces, and at UC Berkeley, where materials scientist Peidong Yang’s lab has built hybrid systems pairing semiconducting nanowires with living bacteria to drive solar-powered chemical reactions. The common thread is diverting the electrons that photosynthesis generates before biology gets a chance to spend them on metabolism, growth, or storage.
The bottleneck was never the light. It was everything the cell does after absorbing it.
That reframing matters because it changes what “efficiency” is even measuring. A plant’s efficiency describes stored chemical energy as a fraction of incoming sunlight. A bio-solar cell’s efficiency describes extracted electrical current as a fraction of incoming sunlight. Those are different accounting systems, and the second one has fewer places for energy to leak out. Skip the Calvin cycle, skip photorespiration, skip the metabolic overhead of keeping a whole organism alive, and the same photon becomes a more direct path to usable energy. Three times the efficiency of a leaf is a meaningful engineering result. It is not a violation of any physical law, because the comparison was never really a fair fight to begin with.
Three Times a Small Number Is Still a Small Number
The context that matters here is what three times a very low baseline actually adds up to. If field photosynthesis runs around 1 percent efficiency, tripling it lands somewhere in the low single digits, well short of what commercial photovoltaics already deliver.
| System | Approximate solar conversion efficiency |
|---|---|
| Field crops (C3, average conditions) | ~1% |
| C4 crops (sugarcane, maize), theoretical max | ~5-6% |
| Biohybrid bio-solar cells (lab reports, this class of device) | ~3-6% |
| Dye-sensitized solar cells | ~11-13% |
| Commercial silicon photovoltaic panels | ~18-22% |
| Laboratory record silicon and perovskite cells | 25%+ |
Silicon photovoltaics are not threatened by this technology, and nobody working in the field is arguing they should be. The pitch for biohybrid solar cells has never been raw efficiency against silicon. It is biodegradability, low-cost fabrication using biological components instead of refined semiconductor materials, and the possibility of integrating power generation directly into living or biologically derived systems, including implantable bioelectronic devices, environmental sensors deployed in places a battery is inconvenient to replace, and hybrid systems that generate electricity while simultaneously performing carbon fixation or bioremediation.
That distinction is also why the Stanford comparison to plants, while accurate, can read as more dramatic than it is. Beating photosynthesis is a low technical bar. Approaching the performance of an ordinary rooftop solar panel is the bar that actually matters for deployment, and biohybrid systems remain well below it.
Durability, Not Peak Output, Is the Real Bottleneck
The harder problem in this field has never been squeezing more current out of a photosynthetic protein complex under laboratory lighting. It has been keeping that protein complex alive and functional outside the cellular environment that evolved to protect it. Photosystem I and Photosystem II are extraordinarily sensitive pieces of molecular machinery. Removed from a living cell and bonded to an electrode, they tend to degrade within hours to days, losing structural integrity and electron transfer capacity as they go.
A photosystem that lasts three hours is a demonstration, not a device.
This is the gap between a headline efficiency number measured under a solar simulator on day one, and a device that could actually sit on a rooftop, a wearable patch, or an implanted sensor and keep producing current for months. Whole-cell approaches, where researchers work with living cyanobacteria or algae rather than isolated proteins, trade some peak efficiency for much longer operational lifetimes, because the organism keeps repairing its own photosynthetic machinery the way it would in nature. Which strategy the Stanford team used, and how long their device kept producing current before performance dropped off, are the details that determine whether this result is a laboratory curiosity or the seed of something that scales. Independent replication and peer-reviewed publication will settle that question in a way a lab demo cannot.
The Comparison That Actually Matters Is Coming Later
The current news cycle around this device is built on a comparison to plants because it is an intuitive, attention-grabbing number: three times better than nature’s own solar panel. But plants were never built to compete on efficiency, and biohybrid solar cells have known that for twenty years of research at places like Vanderbilt and Berkeley without generating much public attention, because tripling a 1 percent baseline was never going to be the number that decided whether this technology matters. The number that will decide it is how many days a device like this keeps working unattended, and how cheaply it can be manufactured at a scale beyond a single postage-stamp prototype on a Stanford lab bench. Until that number exists, the honest headline is not that biology has been beaten. It is that biology was never really trying to win this particular race.
Credit: Karol Majewski on Unsplash