Download as PDF, PPTX
Uploaded byGareth Rowlands
123713AB lecture08
This document summarizes MacMillan's total synthesis of callipeltoside C, which employs organocatalysis and several interesting chemical transformations. The retrosynthesis splits the molecule into three fragments - the macrocyclic lactone core, carbohydrate, and a third segment prepared using organocatalysis. The forward synthesis couples these fragments in a convergent manner, with key steps including a Negishi carbometallation, organocatalytic hydroxylation, Semmelhack reaction to form the tetrahydropyran ring, and glycosidation to join the sugar moiety. The synthesis highlights the utility of retrosynthesis in simplifying complex targets and total synthesis in confirming the structure of natural products.
More Related Content
Similar to 123713AB lecture08
123713AB lecture08
- 1. This lecture coversMacMillan’s synthesis of callipeltoside C, a molecule with potential anti-viral and anti-carcinogenic properties. Other groups (Evans, Trost, Patterson & Panek … all names you should become familiar with) have synthesised other members of this family of compounds. This synthesis employs some of MacMillan’s organocatalysis chemistry as well as a number of other interesting (and in one case challenging) chemical transformations. 1
- 2. So let usstart the retrosynthesis … As always it is important to remove sensitive/reactive functionality as quickly as possible so a C–O disconnection allows the sugar to be removed. This also splits the molecule into two readily prepared fragments, the macrocyclic lactone core and the carbohydrate. To simplify formation of the macrolide, first disconnect the lactone. This removes the macrocycle. 2
- 3. C–C disconnection breaksthe molecule into two fragments. In the forward sense this reaction can be achieved by a simple Grignard reaction or equivalent. We now have three segments to prepare. Hopefully you can see how this has simplified the problem greatly. The whole molecule may look daunting but these smaller sections seem readily achievable. The advantages of breaking a molecule into fragments instead of attempting a long, linear synthesis should be obvious … • Each block can be made simultaneously without effecting the others (so when you drop one in the rotary evaporator bath …) • Less concerns about chemoselectivity • The mathematics of a convergent synthesis favour higher yields than with a linear synthesis. 3
- 4. You need toremember that a good retrosynthesis is a road map to preparing the molecule but it is not a definitive instruction manual. Reality may necessitate changing the order of some steps and a complete rethink … chemistry is easy on paper. It is not always so easy in reality. The bottom half of the molecule can be further dismantled. C=C disconnection halves the molecule once more. The forward equivalent is a standard HWE reaction. The right hand fragment was made during an earlier synthesis by Evans. The left hand fragment will be formed using organocatalysis. 4
- 5. Disconnection of thetetrahydropyran is slightly more tricky (but the synthetic equivalent is elegant). C–O disconnection allows ring opening to give a linear molecule. But the disconnection actually involves a great deal of simplification as it involves C–C and two C–O disconnections. Disconnections such as these are hard to see and it is a matter of experience, a good working knowledge of the chemical literature and an even better knowledge of how to search databases effectively. Practice allows you to spot opportunities and exposes you to more chemistry. 5
- 6. But what agreat simplification that cyclisation was … Now all we are left with is a 1,3-diol. As soon as you see this pattern you should be thinking about the aldol reaction … (there are many other efficient ways of introducing this functionality but until you gain more experience the aldol reaction is a great starting point) 6
- 7. The aldol reactionis a reliable method to prepare β-hydroxyketones or β- hydroxylaldehydes. These in turn offer functionality for either more C–C bond forming reactions (an electrophilic carbonyl group) or can be selectively reduced using Evans chemistry. C–C disconnection removes the propargyl group. Substrate control should allow the reaction to be achieved with high diastereoselectivity (remember Cram Chelation and Felkin-Anh will give different diastereomers). C–C aldol disconnection (1,3-diX disconnection) permits two stereocentres to be controlled … 7
- 8. All we areleft with is the synthesis of the sugar moiety. When looking at a target such as this there are two starting points: 1) an existing carbohydrate (boring) 2) ring-open the hemiacetal and synthesis the open chain form. As there are plenty of hydroxyl groups you might want to consider … … the aldol reaction. Here is one 1,3-diX disconnection. 8
- 9. Here is another1,3-diX disconnection. Thus one retrosynthetic route would be: • C–C disconnection to remove methyl group. • C–O disconnection (not shown) using the open chain form of the carbohydrate. • C–C disconnection aldol reaction installs two carbon atoms with control of stereocentre. 9
- 10. The remaining oxygenatedfragment can be formed from yet another aldol reaction and the dimerisation of this simple aldehyde. Having finished the retrosynthesis lets address the synthesis (as most people find this easier to visualise and we need to known that MacMillan’s plan actually works!) 10
- 11. Starting with thebottom half of the molecule … The first step is a Negishi carbometallation- iodination. This permits the stereospecific addition of a carbon fragment and a metal to an alkyne (or alkene). The addition invariably gives the cis-product. The mechanism of the Negishi carbometallation is complex and almost certainly there are three competing mechanisms. Which one is operating will depend on the aluminum reagent and the solvent (amongst other things). A simplified version is given on the next slide … 11
- 12. There is aninteraction between the zirconocene dichloride and the trimethylaluminium, which creates a highly reactive aluminum species. This forms a π- complex with the alkyne. The nucleohilic alkyne attacks the electron deficient aluminium. Simultaneously the nucleophilic methyl group attacks the polarised alkyne (regioselectivity can be explained by the more stable cation). The simultaneous nature of this addition leads to syn addition and the stereospecificity of the reaction. Once the organoaluminium species has been formed it is an example of simple metal-halogen exchange (effectively transmetallation) to give the iodide with retention of stereochemistry. phew … 12
- 13. With the vinyliodide in place the terminal alcohol was oxidised under standard Swern conditions to give the aldehyde necessary for the … … organocatalytic hydroxylation reaction that will introduce the necessary stereocentre to this fragment. 13
- 14. Proline-catalysed asymmetric hydroxylation occursas outlined in lecture 6. Condensation of the proline and aldehyde results in the formation of an enamine. Hydrogen bonding between the carboxylic acid and nitrosobenzene delivers the electrophile to the top face. Standard functional group manipulation prepares this small fragment for coupling to the rest of the molecule. 1) reduction of the aldehyde to primary alcohol 2) reduction of the O-alkyl hydroxylamine 3) chemoselective protection of the primary (less sterically demanding) alcohol 4) orthogonal protection of the secondary alcohol. 14
- 15. Synthesis of thecentral tetrahydropyran. This involves reagent control catalytic direct aldol reaction as covered in lecture 6. Condensation of the proline with the more reactive, less sterically demanding aldehyde creates the enamine that attacks the chiral aldehyde. It might be interesting for you to work out if this is a case of matched or mis-matched substrate-catalyst control … or to look up what this means! Addition of the organozinc reagent occurs with good diastereoselectivity. The reaction is under … 15
- 16. … substrate controlwith the standard Felkin-Anh (apologies for the spelling mistake). Remember. The largest substituent is perpendicular to the carbonyl group. There are two conformations that fulfil this criterion. The nucleophile then approaches along the Bürgi-Dunitz angle attacking through the conformation that has it passing the smallest substituent. Now the scene is set for the Semmelhack reaction … This is a palladium-mediated reaction that closes the ring while inserting carbon monoxide to furnish the ester above. 16
- 17. The mechanism isa little bit of a nightmare (see what I did there!). The Pd(II) is a π Lewis acid. It activates the alkyne towards nucleophilic attack. The oxygen cyclises onto the alkyne (6-exo-trig for those that remember Baldwin’s guidelines) to give the cyclic enol ether. The carbon monoxide adds to the Pd and then participates in migratory insertion to give the acyl palladium species. The methanol then reacts to give the ester and Pd(0) … 17
- 18. … the Pd(0)is oxidised by the benzoquinone so that it can rejoin the catalytic cycle. The reaction of the enoate does not stop here. It too can be activated by the Lewis acidic Pd(II) and this permits formation of an oxonium species which is trapped by more methanol as the ketal. Lovely reaction … Standard FGI prepare the THP for the subsequent coupling reactions: 1) Orthogonal protection of the secondary alcohol 2) Selective deprotection of the primary alcohol with DDQ (electron acceptor - oxidises the para-methoxybenzyl protecting group and thus cleaves it). 3) Parikh-Doering oxidation (like the Swern oxidation this is an activated DMSO oxidation). 18
- 19. Conversion of thevinyl iodide into a Grignard reagent permits the coupling of the two fragments formed so far. The stereochemistry is an example of Cram chelation control. If you do not believe me you should draw out the reaction for yourself (you should probably do this any way as good practice). 19
- 20. The resulting allylicalcohol is methylated. DDQ deprotection of the para- methoxybenzyl protected alcohol is followed by a second Parikh-Doering oxidation. 1) Horner-Wadsworth-Emmons coupling then joins the last part of the southern hemisphere onto the molecule. The HWE reaction is more reactive than the Wittig reaction and the side product is more readily removed during an aqueous work- up. 2) TBAF removes the silicon protecting group. 3) Barium hydroxide hydrolyses the ester. 20
- 21. Yamaguchi esterification formsthe macrolactone. This reaction involves formation of a highly reactive mixed anhydride, which is then attacked by the DMAP (N,N’-dimethyl-4-aminopyridine). The resulting activated ester is attacked by the alcohol to give the lactone. Unfortunately, under the reaction conditions the ketal undergoes elimination … … luckily it can be reintroduced as the desired hemiacetal (with the correct stereochemistry - anomeric effect and bulky group equatorial) by treatment with a mild acid. The strong acid then removes the silyl protecting group. 21
- 22. And finally weare onto the synthesis of the sugar fragment. Here the MacMillan group had a little trouble. They prepared the reported molecule but found that the nmr did not match the published data. 22
- 23. It turns outthat the stereochemistry of the sugar had been miss-assigned when the molecule was isolated. The correct sugar was actually the enantiomer of the reported structure. This highlights another use of Total Synthesis … structural elucidation. The synthesis of the sugar moiety starts with a proline-catalysed aldol reaction. This is a dimerisation of the TIPS protected aldehyde (TIPS = triisopropylsiyl or iPr3Si). 23
- 24. Substrate controlled (Felkin-Anh)Lewis acid-mediated aldol reaction of the silyl enol ether gives the sugar with excellent diastereoselectivity (but not fantastic yield). Even though there is a Lewis acid present … … it is not an example of Cram chelation control as the TIPS group prevents chelation. Like many sugars the molecule is an equilibrium of the open and closed chain forms. 24
- 25. 1) A mixtureof acetyl chloride and benzyl alcohol generates HCL in situ. This is mild enough to deprotect the primary silyl ether (but leave the less reactive secondary silyl ether alone) and form the acetal at the anomeric position. 2) The alcohol is then converted into a xanthate to allow … 3) The Barton-McCombie radical deoxygenation reaction. 4) Oxidation with Dess-Martin periodinane (DMP) gives the ketone (See oxidation lectures) 1) Grignard addition gives one diastereomer with the nucleophile approaching axially. This is probably a result of the equatorial approach being blocked by the methoxy ether. 2) Hydrogenation removes the benzyl protecting group from the anomeric position. 3) Activation of the anomeric oxygen as Schmidt’s trichloroacetimidate prepares the sugar for glycosidation. 25
- 26. 1) Schmidt glycosidationjoins the two fragments together and then … 2) the final protecting group is removed. The synthesis is fairly convergent and this allows an increased yield compared to some of the earlier syntheses (of callopeltoside A not C). Longest linear sequence is 18 steps an the overall yield is 12% (although this may be based on recovered starting material, some of the supplementary information is a little ambiguous). But, overall it is a neat synthesis that demonstrates the value of organocatalysis. 26