Biosynthetic specialty chemicals

Specialty chemicals are low-volume, high-value molecules sold for what they do rather than by the tonne. Some are now made biologically and some are not, and the choice is not ideological. There are specific structural properties that decide it.

Where chemistry is better

Synthetic chemistry is extremely good at simple, thermodynamically favourable transformations at high concentration: nitration, esterification, hydrogenation, polymerisation. It runs at high substrate loading in organic solvent, so reactors are small for the output, and separation is straightforward because the product is concentrated.

For a small, achiral, structurally simple molecule, chemistry usually wins outright. Fermentation would deliver it in a dilute aqueous broth full of biomass, and the separation would cost more than the synthesis.

Where biology wins

Three structural features flip the comparison.

Stereocentres. A molecule with several chiral centres has many possible stereoisomers, and usually only one is wanted. Chemical synthesis controls stereochemistry with chiral catalysts, auxiliaries or resolution — each adding steps, and resolution discarding at least half the material. Enzymes are chiral catalysts by construction: an active site binds one arrangement and not its mirror image, so selectivity comes free rather than being engineered in. The more stereocentres, the larger the advantage.

Regioselective oxidation of an unactivated C–H bond. Putting a hydroxyl at one specific carbon of a molecule that has many similar carbons is one of the hardest things in synthetic chemistry, normally requiring protecting groups on everything you do not want touched — steps that add nothing to the product and are removed later. Cytochrome P450s do it by holding the substrate in a fixed orientation so only one bond reaches the reactive iron-oxo species. Positional control by geometry rather than by protection is the single largest structural advantage biology has.

Ring systems and long convergent routes. Where a chemical route needs fifteen steps, each with a yield below one, the overall yield collapses multiplicatively. A pathway does the same work in one vessel.

The ceiling nobody escapes

Yield from sugar is bounded by electron bookkeeping. Converting sugar into a product more reduced than sugar — a terpene, a fatty acid, a hydrocarbon — requires reducing equivalents, and generating them means burning some carbon to CO₂. So a highly reduced target has a lower theoretical carbon yield than an oxidised one, before any inefficiency. This sets an upper bound that strain engineering can approach and never exceed, and it is why published yield projections should be checked against pathway stoichiometry rather than taken as ambitions.

Why dilution costs money

Separation cost scales with how much material you must process per unit of product. At 1 g/L, a tonne of product means handling a thousand tonnes of broth; at 100 g/L, ten. Titre therefore pays twice — more product, and less to separate it from — which is why titre, not pathway novelty, is the number that decides whether a route is commercial.

Volatile and hydrophobic products can escape this partly through in-situ removal into a second phase. Water-soluble ones cannot, and for them the separation train usually dominates the cost.