Bacteria Grow Their Own CdS Semiconductors and Triple Chemical Output

Engineered Bacillus subtilis coated with a self-grown cadmium sulfide semiconductor produced nearly three times more 2'-fucosyllactose under illumination than uncoated cells in the dark. That figure, reported October 27, 2026 in Advanced Science, comes from a study on semi-artificial photosynthesis in non-photosynthetic bacteria (DOI: 10.1002/advs.78265).

The work targets a basic accounting problem in fermentation. A bacterium making a valuable chemical draws both its carbon and its energy from sugar in the fermenter. The two demands compete. Sugar burned for energy cannot become product, which caps how efficiently a feedstock converts into output.

Photosynthetic bacteria solve this by drawing on sunlight, but they are poor substitutes for established production hosts whose genetics have already been engineered over many years. The researchers, working with B. subtilis, took the alternative route: keep the proven organism and give it an artificial light-harvesting capability.

How does a bacterium grow a semiconductor?

The team did not synthesize nanoparticles separately and mix them with cells, the usual approach in semiconductor-microbe hybrids. Instead, they exploited the bacterium's own sulfur metabolism. When B. subtilis breaks down the amino acid cysteine, it releases hydrogen sulfide. In a medium containing cadmium ions, that sulfide reacts at the cell surface to form CdS, a semiconductor that absorbs visible light and generates energetic electrons.

The placement matters as much as the chemistry. Because the sulfide originates inside the cell, the semiconductor forms directly on the surface, where generated electrons travel the shortest distance to reach cellular machinery. Electron microscopy showed dense deposits around the bacteria, and X-ray diffraction identified the material as crystalline CdS. Illuminated cells also produced a photocurrent, confirming the coating was photoactive.

The confirmation ran both ways. Disrupting the bacterium's H₂S-producing pathways reduced both hydrogen sulfide output and the amount of semiconductor that formed.

Did the light-generated electrons reach metabolism?

They did. Beyond the near-tripling of 2'-FL output — a sugar used in infant formula — the coated cells contained more NADPH, the reducing power that drives biosynthesis. Light was not merely generating charge at the surface; the extra energy entered metabolism.

The attachment mechanism appears decisive. Commercial CdS nanoparticles simply mixed with the bacteria gave a much smaller boost. Close contact between the inorganic material and the cell wall seems required for efficient electron transfer, and B. subtilis's thick peptidoglycan wall makes loose attachment worse. The exact electron path across the boundary remains unresolved.

The effect also generalized. In a second engineered strain, illumination raised production of surfactin, a chemically unrelated molecule, by about 66%. Light could therefore supply reducing power from outside the cell rather than through repeated rewiring of each biosynthetic pathway.

What stands between the flask and the factory?

Cadmium is toxic. The researchers removed almost all of it during purification, but for a product such as 2'-FL, destined for infant nutrition, that remains a serious complication. All results came from 250 mL flasks; in a dense industrial fermenter, delivering enough light to semiconductor-coated cells becomes a major engineering problem in its own right.

The study's value therefore lies less in cadmium sulfide as a coating than in the demonstration that a cell can build its own electronic interface. Part of the burden of metabolic engineering can, in principle, move outside the genome. Whether the idea survives safer materials and industrial-scale reactors with adequate light delivery is the question that now matters.