[kriggy]

I think the grignard reagent adds from the less hindered side which in this case is the one you want

[phth]

Dan: I like your ideas with the alkene metathesis .  I messed up my synth should convert the acid to ketone before the double bond is dihydroxylated.

(I'm here to learn)
Why do you hydrolize acetyl groups with LiOH? Any reason to choose it instead of NaOH or other option?

Great post!

Thanks!  Doing these types of problems is definitely the best way to learn o-chem!

As for why I chose LiOH, it was mostly because that's what I used in lab!  There are many cases where the difference between Li, Na, and K will give you a change in rates or selectivity.  For instance LiBH₄ reduces ketones with a ten fold rate increase over NaBH₄.  The rational I've heard for this is that the lithium binds more strongly to the carbonyl oxygen, thus activating the ketone during the rate-limiting step, addition of the hydride.  I've heard a similar argument for the hydrolysis of esters with LiOH/MeOH/H₂O.  Supposedly it forms a six-membered transition state to help activate the carbonyl (see picture).  Take this explanation with a huge grain of salt because I can't find any paper that supports this hypothesis or the fact that LiOH does have a faster rate compared to NaOH or KOH.  Maybe some of the more experienced members would know better.

Interestingly when 2,2,2-cryptand or a crown ether is used, the yield drops considerably;  LAH reductions wont go without Li, so a logical conclusion is to substitute Li with a lewis acid and it changes the yield from good to quantitative (i.e. 100%).  FeCl3, CoCl3, etc. do this with LAH; I think this is an example of the elegance in details, and how they can lead to perfection when textbook reasoning says H^-+electrophile will react.  Really reducing agents are not H^-.  Even R-Li compounds are covalently bonded to Li

[Dan]

Hi Dan,

That seems like a nice proposal. I wonder how did you get to the Si-catalyst procedure?

I was aware of this kind of gamma crotylation reaction with boronates and trifluoroborates, so was googling around and found these commercially available (albeit expensive) "EZ-CrotylMix" reagents.

I have got one question about this route though: does the Grignard reaction give a reliable yield/product here?
And by alkylating the ester you create another chiral center, so only one of the two products (si or re product) has the hydroxyl group in the right position to kick out the -OTBS-group?

I might be missing something here, so feel free to correct me!

The Grignard addition should be fine. I used to do this kind of reaction with 5- and 6-membered lactones routinely in the lab.

The stereochemistry of the Grignard addition doesn't matter. The mechanism of the final step is an acid-catalysed desilylation and then acetal formation in the normal way (not an S_N2 reaaction with TBSOH as the leaving group). Since the reaction proceeds via a planar oxonium, the stereochemistry of the hemiketal is inconsequential (see below).

[critzz]

Ah I see, forgot that silyl-protecting groups are pretty acid-labile (because I mostly see them cleaved using F-). Thanks for explaining!

[Dan]

I'm going to come back and try to improve it, but other criticism are welcome.

Could you comment on and/or give a reference for the chemoselectivity of the NaBH₄/HFIP reduction? I've not seen this modification before and would be interested to read more about it.

[SinkingTako]

It's Saturday again. The winner for this week is Dan!

Could Dan kindly post a new problem? I think starting a new thread is better, so that the current discussion can go on as well.

So happy for the active participation everyone, and there's so much to learn!