Researchers in China have built a cell that isn’t alive. A hollow sphere of cadmium sulfide wrapped in polydopamine, no larger than a virus, that performs one of the most sophisticated tricks in biology: it eats sunlight and exhales hydrogen peroxide. Not quite photosynthesis as a leaf does it, but close enough that the distinction starts to feel like pedantry.
The work, published in the Journal of the American Chemical Society two weeks ago by Can Li’s group at the Dalian Institute of Chemical Physics, is part of a quiet race to build artificial cells — synthetic structures that mimic the compartments, relays, and confined microenvironments that make biological chemistry so ruthlessly efficient. Nature has a few billion years of head start. These researchers are trying to close the gap with nanotechnology and a lot of patience.
The nanoreactor they built has two features borrowed directly from living cells. The first is a proton relay: a dynamic redox pair in the polydopamine shell that shuttles protons back and forth, accelerating the electron transfer that drives the reaction. It doesn’t pump protons the way a real cell does — there’s no ATP, no membrane potential, no living machinery — but it mimics the effect, and sometimes the effect is enough. The second feature is the hollow architecture itself: a porous shell surrounding a nanoscale cavity that traps photons, enriches reactants, and creates the kind of confined microenvironment where chemistry behaves differently than it does in a test tube.
Together, these two borrowed tricks achieve something that flat, unstructured catalysts struggle with: they balance the mismatched speeds of water oxidation and oxygen reduction, the two half-reactions that need to dance in sync for artificial photosynthesis to work. Under visible light, the system produces hydrogen peroxide at 3.24 millimoles per gram of catalyst per hour, with a solar-to-chemical conversion efficiency of 1.2%.
One point two percent. That number sounds small, and in some ways it is. A commercial silicon solar cell converts light to electricity at 20% or better. But photosynthesis — real photosynthesis, the kind that feeds the world — operates at roughly 1-2% efficiency in most plants. The nanoreactor is already nudging the lower bound of biological performance. And unlike a leaf, it doesn’t need water, soil, or a growing season. Encased in a sodium alginate hydrogel, it sits in sunlight and keeps working, recyclable, inert, patient.
There’s something quietly heartbreaking about this kind of research. We are building ghosts of cells — hollow, lifeless, meticulously engineered — because the real thing, for all its elegance, isn’t enough. We need more energy than photosynthesis can harvest, more chemistry than biology can provide, more control than evolution ever selected for. So we strip living systems down to their functional principles and rebuild them in dead matter, hoping the copy will outperform the original.
It usually doesn’t, not yet. But the gap is closing. And each iteration teaches us something about why cells work the way they do — why they compartmentalize, why they relay protons, why they confine reactions in tiny spaces where the rules of bulk chemistry bend and warp. The nanoreactor isn’t alive, will never be alive, and isn’t trying to be. But it is, in a very real sense, learning from the dead what the living have always known.
Can Li put it plainly: the work opens opportunities in artificial photosynthesis, energy catalysis, and synthetic chemistry. Which is true, and which is also the kind of thing scientists say when they aren’t quite sure yet what they’ve built. The history of artificial photosynthesis is littered with elegant catalysts that worked beautifully in the lab and dissolved in sunlight, or scaled poorly, or required precious metals that made them economically absurd. This one uses cadmium sulfide and polydopamine — not exactly rare earths, but cadmium carries its own environmental weight.
Still. A hollow sphere that traps light like a leaf, moves protons like a membrane, and produces useful chemistry from nothing but water and sun — that is not nothing. It is a small, synthetic step toward a world where energy doesn’t have to be dug out of the ground or captured by living things that need time and water and luck. A world where we build our own metabolism, our own slow breathing ghosts, and set them in the sun to work.
Whether that world is closer than it was last month, I don’t know. But someone in Dalian just taught a dead shell how to eat light. That seems worth writing down.
Source: “Biomimetic Redox-Mediated Proton Relay in Nanoreactors for Photocatalysis”, Li et al., *Journal of the American Chemical Society, July 8 2026. Additional reporting via SciTechDaily and EurekAlert.*