[curiouscat]

Are there any general heuristics about when a Biochemical Route to a chemical molecule is competitive versus an Organic Synthetic Route?

My naive understanding was that once the number of steps or the complexity of a target increases a synthetic route becomes less favorable because of the multiplier effect of step-wise yields, intermediate isolation etc.

On the other hand, for the smaller "commodity" feed-stocks a chemical route is favorable because the throughput of a chemical catalyst & typical high T conditions is large compared to a bio-route. Especially when the cost of growing and processing biomass in a bio-reactor tends to be high.

Another place where bio-routes win are chiral molecules.

Are there other heuristics or comments that people might have?

Context: I was motivated to ask when reading up on Phenylacetylcarbinol. This molecule looks fairly simple and I was surprised to read that the dominant commercial process is fermentation and not chemical. Well, in hindsight Ethanol is a simple molecule too.

[discodermolide]

There are well established routes for enzymatic racemate resolution, some dependent on alcohol dehydrogenases. In the molecule you cite, you make the acetate of the racemate then enzymatically cleave the (S) enantiomer. The (R) gets re-cycled by base treatment to the racemic alcohol and then repeat.
If you are lucky you will be able to achieve a dynamically controlled resolution. See https://en.wikipedia.org/wiki/Kinetic_resolution

[curiouscat]

Thanks @disco! Very interesting.

So, say, some use-case turned up where  Phenylacetylcarbinol was needed but the optical purity was irrelevant. Do you think industry would then have used a chemical synthetic route instead of finicky fermenters?

What's your guess. And if you had to make  Phenylacetylcarbinol by conventional chemical synthesis what route might you use?

[discodermolide]

If you don't need both the enantiomers then an enzymatic synthesis is still the way to go. See http://onlinelibrary.wiley.com/store/10.1002/bit.260210614/asset/260210614_ftp.pdf?v=1&t=id39kago&s=feebcfbfe640ff4635d9a927f4529ac75b24b30c for yeast fermentation.
or
http://www.google.com/patents/EP0445129A1?cl=en

Conventional synthesis is tricky and about 4 steps, make a dithiane with acetaldehyde then form the anion and acylate with benzaldehyde, remove the dithiane.

[curiouscat]

Conventional synthesis is tricky and about 4 steps, make a dithiane with acetaldehyde then form the anion and acylate with benzaldehyde, remove the dithiane.

So non-intuitive. Because so many molecules with a very similar backbone (at least to my naive eyes) to Phenylacetylcarbinol seem to be all manufactured synthetically. No mention of a dominant fermentation route.

e.g. Phenylglyoxylic acid, Mandelic acid, Mandelonitrile, 2-PhenylEthyl Alcohol, 1-PhenylethylAlcohol, EthylBenzene, Acetophenone

Perhaps my notion of "similar" is distorted or immature. It seems so unexpected to me that so similar a analog would have a tricky synthetic route.

[discodermolide]

Well sure they are all similar. But take the mandelic acids, you still have to convert the acid to a ketone. Or the gyloxylic acids, acid to ketone then reduce or reduce then acid to ketone.
Similar problems with functionalisation in the other molecules, so not that trivial, especially on >100 ton a month as you make!