[ghovinsen]

Previously related thread: http://www.chemicalforums.com/index.php?topic=87947.0

Continuing my study with the synthesis of the neuraminidase inhibitor Tamiflu, I encounter some reaction scheme I couldn't wrap my head around.


From my knowledge, I deduce:

17  18 is MOM deprotection by acidic condition.
18  19 is an Sn2 type reaction in which the amine lone pair attacks the quaternary carbon of TrCl.

• Is the role of Et₃N to replace the Cl atom prior to Sn2 to increase its leaving group ability?
• Amines are more basic, thus only the amine participates in Sn2 reaction.

19  20 is an Sn2 type reaction in which the amine lone pair attacks the adjacent carbon atom

• Does Et₃N work the same way as above to replace the Cl atom of MeSO₂Cl
• The hydroxyl group is mesylated, increasing its leaving group ability. Sn2 reaction proceeds to furnish an aziridine.

20  21 Aziridine ring opening in which the oxygen lone pair of alcohol attacks selectively at the C6 position.

• BF₃·Et₂O is known to be a catalyst for aziridine ring opening, but in this scheme 1.5 eq was added. What is the role of this Lewis acid?
• Why attack at the C6 position occurs much more selectively?

21  22 Acetylation at of the amine

• Acetylation of alcohol with pyridine as the solvent and DMAP as catalyst is well known. I couldn't however, figure out a reaction mechanism for it.
• How is the triphenyl group removed this way?

This is my best guess for the reaction mechanism:


22  23 Staudinger reduction of the azide

• In the previous steps prior to compound 17, the author used Lindlar catalyst and later KOH and THF/H₂O to reduce the azide and at the same time saponify the ester into a carboxylic acid.
• The Staudinger reduction accomplishes the same, but how is the ester not saponified under basic condition?

My best guess for the reaction mechanism:


Sorry for the lengthy post! I hope I could get some nice discussion out of it.

Thanks!

[TheUnassuming]

Lets break this post down a little and start with the first three:

18->19:  What have you learned about SN2's on hindered quaternary carbons?

19->20:  I forget the reference for this, but NEt3 does something interesting with MsCl.  What is the most acidic proton in the system?

20->21: Maybe drawing this in 3D (or using models) will give you a good answer to the selectivity.  The other possible reason comes from sterics. 

[Dan]

18  19 is an Sn2 type reaction in which the amine lone pair attacks the quaternary carbon of TrCl.

Why S_N2 and not S_N1?

• Is the role of Et₃N to replace the Cl atom prior to Sn2 to increase its leaving group ability?

Triethylamine is generally not considered a nucleophilic catalyst as far as I'm aware. What byproduct would be produced if triethylamine was not present, and why might this cause problems?

• Does Et₃N work the same way as above to replace the Cl atom of MeSO₂Cl

This one is quite interesting actually, see post above.

• BF₃·Et₂O is known to be a catalyst for aziridine ring opening, but in this scheme 1.5 eq was added. What is the role of this Lewis acid?

Are you aware of the concept of "product inhibition" in catalysis? Do you think it could come into play here?

• Why attack at the C6 position occurs much more selectively?

You will need to do a conformational analysis to rationalise this. Most textbooks cover this in the context of epoxide opening, maybe see what you can find on that for ideas.

• How is the triphenyl group removed this way?

The byproduct in this reaction is pyridinium acetate, what is its pKa? Could it catalyse something?

• In the previous steps prior to compound 17, the author used Lindlar catalyst and later KOH and THF/H₂O to reduce the azide and at the same time saponify the ester into a carboxylic acid.
• The Staudinger reduction accomplishes the same, but how is the ester not saponified under basic condition?

I agree, this is confusing, could it be a misprint? Can you give the citation information for paper so we can have a look?
[/list][/list][/list]

[ghovinsen]

Thank you for the suggestions!

18  19 I missed this simple mistake. Should be Sn1 instead of Sn2. And I guess Et₃N is used to remove the resulting hydrochloric acid.

19  20 Same with above but to remove methanosulfonic acid.

20  21 I've heard of product inhibition in the context of enzymes, but not in organic chemistry.

21  22 All I know so far is that pyridine can work in the same way as DMAP, where the lone pair of nitrogen "activates" the acetic anhydride.

Reference for this paper is DOI: 10.1021/ja963036t (http://pubs.acs.org/doi/abs/10.1021/ja963036t)

[TheUnassuming]

19  20 Same with above but to remove methanosulfonic acid.

Not quite, check my previous post.

20  21 I've heard of product inhibition in the context of enzymes, but not in organic chemistry.

Another way to think/look at this is: what else could BF3 coordinate with that might "use up" your catalyst.

21->22: The activation with pyridine goes generally the same as DMAP, though DMAP is a better catalyst because it tends to favor a different form when its activated an acyl group. 

For your mechanism, try letting the acylation take place fully to form the positively charged and Tr protected nitrogen amide.  What do you know about Tr-N deprotection?

[Dan]

Reference for this paper is DOI: 10.1021/ja963036t (http://pubs.acs.org/doi/abs/10.1021/ja963036t)

Check Scheme 4 - compound 6 is the carboxylic acid not the methyl ester (so KOH is for saponification).

[Dan]

Actually, for the reaction sequence 20  22 the authors don't specify at which point in the sequence the Trt is cleaved. Do you think it is more likely to cleave in the presence of BF₃/ROH (at 75 °C), or with Ac₂O/Py?

[ghovinsen]

19  20 Ah right, the HCl formed then.

20  22

Another way to think/look at this is: what else could BF3 coordinate with that might "use up" your catalyst.

I'm not sure what the catalyst in this reaction is. But I guess the BF3 could coordinate with the negatively charged azide nitrogen, therefore blocking the approach of alcohol to C5. BF3 could also coordinate with the carbonyl oxygen, but that is too far away.

For your mechanism, try letting the acylation take place fully to form the positively charged and Tr protected nitrogen amide.  What do you know about Tr-N deprotection?

Actually, for the reaction sequence 20  22 the authors don't specify at which point in the sequence the Trt is cleaved. Do you think it is more likely to cleave in the presence of BF₃/ROH (at 75 °C), or with Ac₂O/Py?

According to Greene's Protective Groups in Organic Chemistry (4th Ed.), trityl protecting groups for alcohols can be cleaved with BF₃·Et2O either with the conditions:

• BF₃·Et2O, HSCH₂CH₂SH (80% Yield)
• BF₃·Et2O, DCM, MeOH, r.t. (80% Yield)

In this case, conditions of BF₃·Et2O/ROH should allow for aziridine ring opening while at the same time removing the trityl-amine protecting group. If what I said about BF₃ coordinating with azide to block nucleophile access to C5 is true, then formation of ether at C6 could be selectively done. Then the next reaction with Ac₂O/Py system simply introduces an acyl group to the amine.

22  6

Reference for this paper is DOI: 10.1021/ja963036t (http://pubs.acs.org/doi/abs/10.1021/ja963036t)

Check Scheme 4 - compound 6 is the carboxylic acid not the methyl ester (so KOH is for saponification).

Oh I must've missed that!