Total Synthesis of (+)-Roxaticin via C-C Bond Forming Transfer
Hydrogenation: A Departure from Stoichiometric Chiral Reagents, Auxiliaries,
and Premetalated Nucleophiles in Polyketide Construction
Soo Bong Han, Abbas Hassan, In Su Kim, and Michael J. Krische
Additional information (DOI’s):
This time some chemistry for all the transition metal fans out there: a very nice synthesis of (+)-Roxaticin from Krische et al. which demonstrates the outstanding potential of their stereoselective allylation chemistry. By employing the strategy developed by the Krische lab the otherwise painful synthesis of this beastie was extremely simplified. If you are interested in syntheses from other labs have a look in the supporting information of the original paper which contains an useful overview.
Or if you are equally enthusiastic about the chemistry you should have a look in the paper which I linked above under additional information.
So let’s have a brief look at our target:
Obviously the molecule is perfectly suited with respect to the allylation chemistry which was employed. Before I get started with pointing out the individual steps first the syntheses of the main catalysts used:
The first one was synthesized by mixing [Ir(cod)Cl]2, the BIPHEP ligand, chloro-nitro-benzoic acid and the base in the presence of allyl acetate. This in situ formed catalyst was used as such with remarkable results.
The second one was synthesized in a similar manner:
As you will see in the ongoing synthesis the enantiomer of the (R)-I-Cat. was used, too.
The synthesis begins with propanediol which was converted to the bishomoallyl alcohol under standard conditions in good yield and extremely high ee and dr. Protection as the acetonide was followed by ozonolysis and reductive work-up to give the next diol. This was again converted into the bishomoallyl alcohol, protected as the TBS ether and reacted with ozone/sodium borohydride to give another diol. Allylation, conversion of the TBS ether into the acetonide and ozonolysis/reduction furnished the last diol.
In only nine steps the whole “alcoholic”-part of the target was finished! Nice…
Next the C2-symmetric molecule was selectively converted on one side into the alkene by employing some Mukaiyama chemistry. First one alcohol was converted into a selenide which was oxidized and eliminated to give the terminal alkene in moderate yield. Olefin metathesis with the PMB protected homoallyl alcohol shown was followed by another allylation step of the remaining alcohol with methylallyl acetate. This time the second catalyst (S)-II was used and not less than two stereogenic centers were set up.
Another olefin metathesis was employed again with Grubbs-Hoveyda-II but this time with acrolein as the chain extension. Protection of the free alcohol as the TES ether and oxidative PMB removal produced again a homoallyl alcohol. HWE-reaction of the terminal aldehyde and saponification of the ester then furnished protected (+)-Roxaticin in its open form.
Yamaguchi ring macrolactonization and global deprotection with DOWEX-50 finished the synthesis in only 20 steps in the longest linear sequence.
For all those who are interested in the mechanism and have to time to look in the reference, here is the mechanism of this cool allylation step:
The in situ formed catalyst first oxidizes the alcohol to the aldehyde and forms a hydrido-iridium-species. It should be noted that the reaction can also be done with the aldehyde oxidation level but in this case they were too unstable to be used.
Next fresh allyl acetate reacts with the reduced form of the catalyst to give again the allyl coordinated iridium which in turn inserts itself into the double bond an in situ formed aldehyde. By reacting with another molecule of alcohol the homoallyl alcohol was set free and the reaction cycle goes on. For clearance I skipped some intermediates but this should be sufficient to get a brief overview.
I really like this sort of chemistry. Seems not applicable for multigram synthesis but for quick access to a lot of analogues it is the perfect method I think.