A composite grown by three bacteria matched engineering plastic on every mechanical measure and absorbed a hundred and fifty times its own weight in water, so I spent a thesis on the water.
Biomineralised bacterial cellulose is grown rather than moulded: three fermentations lay down cellulose fibres wrapped around carbonate crystals, giving a material with the toughness of nacre and properties comparable to the ABS plastic used for casings and furniture. My thesis took on the single property keeping it off the market, and tested three bio-based coatings against it.
My supervisors were working on the composite described by Yu and colleagues in 2021 and needed its water resistance characterised and improved, using treatments simple enough not to undo the point of growing a material in the first place.
The composite is forty per cent bacterial cellulose, and bacterial cellulose is one of the most water-hungry substances available. Soaked, it swells and the mechanical properties that made it interesting disappear. Everything attractive about the material was intact; one property decided whether it could leave the laboratory.
Three bio-based coatings tested across concentrations, application methods and three levels of wetness, with a custom abrasion rig built on a 3D printer. Stearate took the surface to 136° and held a droplet long enough for it to evaporate; chitosan cut water uptake by forty per cent; and the two findings belong to different coatings, which is the useful part.
Grown composites replace moulded plastic in furniture and casings, because they survive a wet room.
The composite is made by three organisms working in sequence. One lays down a dense mat of cellulose fibres, another contributes a polymer that organises the mineral phase, and carbonate crystals of calcium and magnesium grow through the mat. The result is sixty per cent mineral by mass, with the fibres wrapped along and around the crystals, which is close to how nacre is built inside a shell and is why the mechanical numbers land where they do.
Hard and tough at the same time is unusual for a bio-based material. Natural materials tend to offer one or the other, and the combination is most of why synthetic plastics are still specified for things that have to survive being handled. This composite has the combination, and it is grown at room temperature in a vat.
Bacterial cellulose absorbs up to a hundred and fifty times its mass in water. In the composite that means a sample left in contact with water swells, softens and loses the toughness that justified growing it. The original paper proposes furniture and protective garments, and both of those meet water in ordinary use.
Keeping water off a surface works in two ways, and a lotus leaf uses both. Its waxes lower the surface energy so that water molecules prefer each other to the leaf, and the microscopic bumps covering it mean a droplet cannot touch the surface properly without being pushed. A coating can borrow either mechanism. I was looking for treatments that use one of them, come from biological sources, and can be applied by dipping a dried sheet, because any method needing specialist equipment would cancel out the advantage of a material that grows in a tank.
The material already performed like engineering plastic. What kept it in the laboratory was that it drank.
Potassium stearate was the interesting one, because it does not sit on top of the material. Stearate ions exchange with the carbonate crystals already in the composite, so the mineral phase that gives the material its toughness also anchors the water-repellent layer. Two pieces of evidence support that. Fifteen minutes in the solution beat a three-second dip at every concentration, which is what a reaction looks like and not a coat. And the same treatment on plain bacterial cellulose, with no minerals in it, performed significantly worse than on the composite.
Chitosan, made from the shells that are the second most abundant source of polysaccharide after cellulose, and zein, a maize protein that is more than half hydrophobic amino acids, both behave differently depending on how much you use. Thin, they leave microscopic structures on the surface. Thick, they dry into a continuous film. Under the microscope the zein structures turned out to be flowers, four to seven micrometres across with petals of three to six, which I could not find described anywhere in the literature.
A coating can be tested at three levels of severity, and the three do not have to agree. The first is the angle a droplet holds on the surface, where anything above ninety degrees counts as water-repellent. Uncoated, the composite sits at fifteen degrees. Fifteen minutes in stearate cleared ninety at every concentration tested and reached a hundred and thirty-six.
The second is how long a droplet survives before it soaks through. Stearate held one for seventy-two and a half minutes against a calculated evaporation time of sixty-nine, which means the droplet never entered the material; it dried on top.
The third is full immersion, and here the ranking inverts. Stearate, which won both surface tests, produced no significant improvement in how much water the sample took up. Chitosan at one per cent, which barely passed the first test, cut water uptake by forty per cent at fifteen minutes. A grafted monolayer changes what the surface does to a droplet and does nothing about a sheet sitting in a bath; a film is a barrier. Whichever one you want depends on whether the product gets rained on or washed up.
A finish on a chair or a case has to survive handling, so every coating was abraded and measured again. There was no instrument for this in the laboratory, so I built one from a 3D printer, using its motion system to drag a weighted sample across a known grit at a fixed force and speed, following a published method so the numbers would mean something outside my bench.
Abrasion sorted the three coatings faster than any other test. The films cracked and the chitosan surfaces ended up wetting more readily than bare material. The zein microstructures were simply scraped away. Only stearate at the highest concentration still cleared ninety degrees afterwards, which is what settles the recommendation: for furniture, a coating that fails the first time something is dragged across it is not a coating.
Drag the drawing sideways to see all of it
The reason none of the three performed better is upstream of all of them. The protocol for casting the composite is not designed to produce flat sheets, and the dried samples come out uneven and porous enough that sanding does not correct them. An uneven surface cracks a film and leaves a grafted layer patchy, and water finds those gaps. Any further work on coatings is premature until the casting step produces a flat sheet, which makes the most useful finding in the thesis a criticism of the step before the one I was asked to work on.
I chose this thesis because the interesting part was already proven. Somebody had grown a material with the mechanical properties of a plastic, and between that result and a product sat one property, measurable in an afternoon with a camera and a syringe. Working on that property taught me more about how materials reach the market than inventing another composite would have.
It is the shape of question I have worked on since, in mycelium textiles and in living building materials. The literature is full of materials that perform. What decides whether any of them get manufactured is usually a dull-sounding property, found late, that nobody published because it is not novel. Finding that property early is worth more to a company than another percentage point of tensile strength.