Specialty & fine chemicals
Thirty-six subjects, from fermenter foam to foundry sand — held together by two patterns: the product that poisons its own production, and the mechanism that cannot reach its target.
This cluster runs from engineered yeast to marine gearboxes, and the subjects look like they belong to different industries. Read together, two patterns recur often enough to be worth carrying into any page here.
The product usually limits its own production. In most biosynthesis the interesting ceiling is not the pathway — it is what the accumulating product does to the system making it. Biosurfactants foam their own aerated fermenter, so the more the organism succeeds the worse its own gas transfer becomes. Capsaicin is cytotoxic to the organism engineered to produce it. Hyaluronic acid raises broth viscosity as titre rises until oxygen can no longer be delivered, so the titre ceiling is set by rheology rather than by metabolism. The same shape appears outside fermentation: a detergent protease digests the other enzymes in its own box, and a foundry binder must be strong during pouring and weak an hour later, so its own thermal destruction is a design requirement rather than a failure.
Once you look for it, this is the question worth asking on any page in the cluster: what does making more of this do to the thing that makes it?
A mechanism is only as good as its delivery. The cosmetics-side subjects share a structure. There is a real, specific, checkable mechanism — a SNARE-complex interference, a photolyase that reverses a pyrimidine dimer in one step, a bacterial β-lyase releasing a thiol, an ecology with a measurable pH — and then a barrier, a residence time or a colonisation problem decides whether any of it reaches the target. The stratum corneum’s roughly 500-dalton limit disqualifies most growth factors on arrival. Skin resists colonisation by transient organisms, which is why postbiotics are the more tractable claim than probiotics. A repellent must evaporate to work, so its duration and its efficacy cannot be optimised separately. Where a page here is sceptical, it is almost always the transport step it is sceptical about, not the biology.
A third thing is worth stating once. “Bio-based” is often a claim about the feedstock, not the molecule. Fermented ethanol and petrochemical ethanol are identical; so are fermented and shark-derived squalane. That claim is legitimate on sourcing and life-cycle grounds and becomes misleading only when it is presented as efficacy or gentleness.
Start with biosynthetic specialty chemicals, which sets out why a fermentation route wins on stereochemistry and complexity rather than on cost, and why recovery from a dilute aqueous broth is where the economics usually decide.
Pages are ordered from the platform chemistry through specific biosynthetic molecules, recombinant proteins and biomaterials, enzymes in formulated products, the skin and microbiome subjects, and finally the industrial performance chemicals.
Articles 37
| Biosynthetic specialty chemicals Biology wins where the molecule is hard to build and loses where it is easy — and the dividing line is stereochemistry and unactivated bonds. |
| Biosurfactants You are making a foaming agent inside an aerated, stirred tank — the process fights its own product. |
| Bio-based solvents 'Like dissolves like' is quantitative — and matching a solvent means matching three numbers at once, not just finding a bio-based liquid. |
| Microbial pigments & bio-dyes Being coloured is easy. Being a dye means staying on the fibre — and that is a bonding problem, not a colour problem. |
| Biosynthetic hyaluronic acid Molecular weight is the product — and the fermentation is limited by the fact that its own product makes the broth too thick to aerate. |
| Fermented palm oil alternatives Palm is the highest-yielding oil crop there is — so any replacement has to beat it on land, not merely avoid it. |
| Biosynthetic cannabinoids One branch point makes every cannabinoid — and changing the starter unit changes the whole family at once. |
| Biosynthetic menthol Menthol does not cool anything. It opens the receptor that reports cold — and only one of its eight stereoisomers does it well. |
| Biosynthetic musks Musk odour lives in a narrow window of ring size — and closing a ring that large is a problem of probability, not of chemistry. |
| Biosynthetic capsaicin Pungency is the pain receptor for heat being opened by a molecule — which is why chilli burns without being hot, and why the producing cell suffers too. |
| Cocoa-free chocolate Chocolate is two independent problems — a flavour built in two stages, and a fat that must crystallise in one particular form out of six. |
| Biosynthetic theobromine Theobromine is caffeine minus one methyl group — and the dog toxicity everyone knows about is a difference in how fast it is removed, not in how toxic it is. |
| Biosynthetic L-theanine An amino acid that is not in any protein, built by an enzyme running a reaction it was not named for — and it reaches the brain by competing for a transporter. |
| Bakuchiol It is not a retinoid — no ring, no polyene, no shared chemistry — yet it moves a strikingly similar set of genes. |
| Perfume as a physical system A perfume is not a mixture that smells of something — it is a mixture whose composition changes continuously as it evaporates. |
| Recombinant human serum albumin Albumin does two unrelated jobs — it holds water inside blood vessels, and it carries almost everything — and it is dosed in grams, which is what makes purity hard. |
| Recombinant botulinum toxin A protease that cuts one protein inside one kind of nerve ending — and the reason 'complex-free' matters is immunological, not chemical. |
| Silk fibroin biomaterials The immune response people associate with silk is to the glue, not the fibre — and once that is removed, degradation rate becomes something you can dial. |
| Cosmetic peptides Almost every question about a topical peptide reduces to one: can a molecule that size cross a barrier built to exclude it? |
| Laundry enzymes The protease in the box is trying to digest the other enzymes in the box — and solving that is most of the formulation. |
| Enzyme and probiotic deodorants Fresh sweat has no smell. The odour is made later, by bacterial enzymes cutting odourless precursors — and that reaction is the only thing a deodorant of this class is trying to stop. |
| Microbiome cosmetics A cosmetic must be preserved against microbial growth. A microbiome cosmetic must not kill microbes. That contradiction is the category's central formulation problem, and it is not rhetorical. |
| Probiotic oral care Caries is not an infection to be sterilised; it is an ecological shift toward acid-tolerant bacteria. That framing is what makes seeding a competing organism a coherent idea rather than a wellness slogan. |
| Anti-dandruff, microbiome-based Malassezia cannot make its own fatty acids, so it digests your sebum to get them — and it is what it leaves behind, not the fungus itself, that irritates the scalp. |
| Fermentation-based hair care The hair shaft is dead keratin with no metabolism. Anything a shampoo actually does biologically, it does to the scalp — everything else is surface chemistry. |
| Antiperspirants without aluminium An antiperspirant plugs a duct; a deodorant changes a chemistry. Dropping the aluminium salt does not make a gentler antiperspirant — it changes which of the two things the product is. |
| Plant-derived antimicrobial hygiene Ethanol from sugarcane and ethanol from ethylene are the same molecule. The feedstock claim is real; it is just not a claim about how the product kills anything. |
| Biological sunscreens and DNA-repair enzymes Placental mammals lost the enzyme that reverses UV damage directly. Putting photolyase in a cream is an attempt to give it back — and the whole question is whether a 50-kilodalton protein reaches a cell nucleus. |
| DNA-personalised skincare The premise is testable and the panels are small. What has to be shown is not that the variants exist, but that knowing them changes the right answer — and for most of the panel the answer was sunscreen anyway. |
| Fermented cosmetic actives and nutricosmetics Eaten collagen does not arrive at the skin as collagen. If it does anything, it is because two- and three-residue fragments survive digestion and act as a signal — which is a different claim, and a testable one. |
| Biodegradable exfoliating microbeads A scrub particle has to be hard enough to abrade dead cells and soft enough not to tear living ones. Polyethylene was good at that, which is why replacing it is a materials problem and not a substitution. |
| Tattoo aftercare, plant-based Almost all the healing benefit of an aftercare balm comes from the wax and oil base holding water in the wound. The botanical actives are the part most likely to cause a problem, because the barrier is open. |
| Plant-based insect repellents A repellent has to evaporate to be smelled, and evaporating is how it runs out. Protection duration is set by vapour pressure — which is why the essential-oil actives are short-lived by chemistry, not by formulation failure. |
| Bio-based corrosion inhibition An underdosed anodic inhibitor is worse than no inhibitor at all — it concentrates the same total attack into a few pits. That asymmetry is why substituting a corrosion chemistry is not like substituting a solvent. |
| Bio-based metal surface treatment No organic coating is impermeable — water and oxygen get through all of them. A coating protects by staying stuck to wet metal, and that is a different property from being a barrier. |
| Biodegradable aviation and marine lubricants The ester bond is what makes these oils biodegradable and also what makes them vulnerable to the water they are sold to be used in. Everything in the category is an answer to that single conflict. |
| Bio-based foundry binders A mould binder has to be strong while molten metal is poured against it and weak an hour later. It is an adhesive designed to destroy itself on schedule, and that is the requirement every replacement chemistry struggles with. |