# retrosynthesis gamma-hydroxy ketone

One of the sections about retrosynthesis in Warren's book addresses how to access $$\gamma$$-hydroxy carbonyl compounds like ZM181. This is again one of the examples to advice the reader to check the results for plausibility: while epoxides often may be a suitable building block, here, the intermediate would react further with the ester present to yield a lactone (second line) instead:

In the edition accessible to mine, Warren apparently does not detail out if an alternative approach with chloro- or bromoethanol (third line) to yield ZM181 were viable.

Question: Would be the approach of the third line, chloroethanol and ethyl acetoacetate to yield ZM181, still viable outside academia, too?

I'm aware the intramolecular transesterification to yield a five-membered cycle needs to be considered as a possible product for both approaches. On the other hand, I think either route could be favoured by choice of the reaction conditions, i.e.

• complete single deprotonation of ethyl acetylacetonate in dry THF by one equivalent $$\ce{NaH}$$ (not two), then followed by the addition of the expoxide may ease the formation of the lactone. The reasoning of $$\ce{NaH}$$ is to ensure all of the starting material is deprotonated once with the irreversible escape of $$\ce{H2}$$. Or,

• the partial, equilibrium based deprotonation of the $$\beta$$-dicarbonyl compound by $$\ce{K2CO3}$$ in a mixture of THF/water, in presence of the bromo ethanol already present, might reduce the concentration of alcoholate and thus lower chances of the formation of the lactone. This line of thought accounts for the different $$pK_a$$ of the reagents present, e.g., water could protonate free alcoholate, too.

Reference: Stuart Warren Organische Retrosynthese, Teubner, Stuttgart 1997.

• I don’t think the second and third lines are realistically different - you’d need a base present in both so if the lactone is favourable , it would form in both cases. Practically, depending on what the next step in the synthesis is, the lactone might be fine (eg can still add a nucleophile in, or reduce it down) Mar 22 '21 at 18:45
• Also the 1,3-dicarbonyl likely exists (to a large extent) as the enolised form, which may protect the lactone from forming immediately after the addition. Mar 22 '21 at 18:46
• The only advantage I see of the bromoethanol route is that you could put a protecting group on the -OH and prevent lactone formation that way Mar 22 '21 at 23:31
• For a moment, I thought running the reaction of the second line in dry THF would indeed yield the lactone. Contrasting to this -- given the different pKa of the beta-dicarbonyl compound and of an alcohol -- I think now, running the reaction of the third line with $\ce{K2CO3}$ in a mixture of water and THF could reduce a the concentration of free alcoholate to trigger the intramolecular transesterification. But still it could be an uphill battle (already because intramolecular reactions are favoured over intermolecular ones, and the formation of a cycle of five atoms, etc.). Mar 23 '21 at 6:27
• What is the green-chemistry aspect of this? The bromide looks hazardous enough to me to kick this out of the category. Mar 27 '21 at 0:24