Vitamin production

Vitamins are not a chemical family — they are a functional category, and each has an industrial route shaped by its own chemistry. Four cases show the range.

B12: nothing else can make it

Cobalamin is among the most structurally complex non-polymeric natural products known: a corrin ring built around a cobalt ion, assembled through roughly thirty enzymatic steps. It is synthesised only by certain bacteria and archaea. No plant, fungus or animal makes it; animals obtain it from microbial synthesis, whether in their own gut, in a rumen, or through the food chain.

That fact settles two things. Industrially, there is no viable total chemical synthesis — the celebrated laboratory synthesis of the molecule took more than a decade and dozens of chemists — so commercial B12 is made by fermentation, typically with Pseudomonas denitrificans or propionibacteria. Nutritionally, it is why B12 is the one nutrient that a diet excluding all animal products cannot supply reliably, and why fortification or supplementation is a requirement rather than a precaution. Claims that particular plant foods, algae or fermented products supply it should be treated sceptically: several contain corrinoid analogues that are detected by some assays but are not biologically active in humans, and may even compete with genuine B12.

Riboflavin: fermentation displaced chemistry

Vitamin B2 was long made by multi-step chemical synthesis. It is now made almost entirely by fermentation, using natural overproducers such as the fungus Ashbya gossypii or engineered Bacillus subtilis.

The displacement was economic and environmental rather than regulatory: the fermentation route uses fewer steps, avoids several solvents and reagents, and produces less waste per kilogram. It is a clear, well-documented example of a biological process outcompeting an established chemical one on cost, which is less common than the field’s promotional literature implies.

Vitamin C: a hybrid, not a fermentation

Ascorbic acid production is a useful corrective to the idea that processes are either chemical or biological. The classical Reichstein process converts glucose to sorbitol chemically, uses a bacterial oxidation to make sorbose — one biological step embedded in a chemical sequence — then proceeds chemically.

The route that displaced it is a two-step fermentation producing 2-keto-L-gulonic acid, which is then converted chemically to ascorbic acid. The modern process is still a hybrid; what changed is how much of the sequence the microbes do. Neither route is a single-organism fermentation from sugar to vitamin.

Vitamin D3: a photochemical reaction, not a synthesis

Vitamin D3 is not made by fermentation at all. 7-dehydrocholesterol — obtained industrially from lanolin — is irradiated with ultraviolet B light. The UVB photon opens the B ring of the steroid, producing previtamin D3, which then rearranges thermally to cholecalciferol over hours.

This is the same two-stage sequence, photochemical then thermal, that occurs in skin, which is why vitamin D synthesis in the body depends on UVB specifically rather than on sunlight generally, and why glass, sunscreen, latitude and season all affect it. Vitamin D2, ergocalciferol, is made by the same irradiation applied to ergosterol from yeast or fungi, and is the vegan alternative; D2 and D3 are not equivalent in potency at raising and maintaining circulating 25-hydroxyvitamin D, with D3 generally the more effective.