Note: I'm currently on holiday. I do have internet access, but drawing chemical structures on the netbook I've borrowed may just be too painful. I'll try and get some updates out ASAP.
Di-t-butylisobutylsilyl, Another Useful Protecting Group
I suspect that anybody who’s been engaged in synthetic chemistry for more than a year or two has probably used a silicon protecting group. I’ve used plenty, and they’re generally very useful, easy to put on and take off, and pretty robust under a lot of different conditions. One of the great things about these groups is the huge range available; from the labile TMS and TES, to the more robust (and useful) TBS and TBDPS, to the hardy TIPS. At the extreme end of the scale an even tougher group is the tri-t-butylsilyl group, but that’s very hard to put on or take off, and the silylating reagent itself is a pain to make. This week Corey published an attempt to fill the niche for a group tougher than TIPS, but more useful than tri-t-butylsilyl, with the disclosure of the di-t-butylisobutylsilyl (BIBS) group.
As far as I'm aware, the first silicon protecting group which was deliberately designed was the TBS group, introduced to the world by the Corey group way back in 1972. This paper also disclosed a number of important protocols still in widespread use, such the use of TBAF for the removal of TBS specifically, and silyl groups in general, as well as the use of imidazole (and hence N-silylimidazoles) in silylations. I even learned something on rereading the paper for this post:
We have also made the interesting and useful observation that trimethylsilyl and triphenylsilyl ethers can be hydrogenolized easily to form alcohols (e.g., using 10% Pd/C, 1 atm H2, ethanol solvent, at 25°C for 1-10 hr), and further, that dimethylisopropylsilyl ethers undergo slow but appreciable hydrogenolysis under these conditions.
Actually, I’m pretty sure that just stirring many TMS ethers in ethanol would result in deprotection - I've never even been able to column any I've made. Having said that, I’ll never forget how a tertiary TMS ether survived 15 steps (including 49% HF to remove a primary TBS ether) in Wender’s total synthesis of phorbol.
BIBSOTf is, according to a note in the paper acknowledgements, available commercially from Gelest, Inc.. That's all well and good for American chemists, but if you live in the UK or the rest of the world you’ll probably have to make it for the forseeable future. A long time ago I remember reading in Kocienski’s excellent book on protecting groups that TBSCl can (and should) be freshly prepared from dichlorodimethylsilane and t-BuLi and thinking, ‘Eh? Why? Who has the time?’. I mean, what would your supervisor say if they caught you making TBSCl? And t-BuLi, is, well, not something you use if you can help it. BIBSOTf is prepared by reacting commercially available isobutyltrichlorosilane with three equivalents of t-BuLi, giving first BIBSCl, then, after heating to 120°C, BIBSH by hydride transfer from the third equivalent. That’s right, they use t-BuLi as a reducing agent. This BIBSH can then be converted to the desired BIBSOTf by treatment with TfOH (accompanied by evolution of H2; if anyone knows how this works I’d love to hear from you!) in a respectable 70% yield over the two steps on 20 grams. So, assuming you’ve survived the process, what can you do with your nice freshly prepared BIBOTf?
Well, you can protect phenols, but not all that easily; phenols with para electron withdrawing groups are okay, but phenol itself is pretty slow. The resulting BIBS ethers, although ‘readily cleaved’ with TBAF are around 1300 times more base stable than TIPS and 10000 times more base stable than TBS. Owing to their enormous bulk, the group is quite useful for ‘directing’ electrophilic aromatic substitution; bromination of PhOBIBS only occurs para, for instance, and you may remember the cationic cascade that the Corey group published last month, which also made use of this property:
Alcohols can also be protected after prolonged heating, but more interestingly, the group also produces stable BIBS esters from carboxylic acids, and can even be put on primary amines. The bulk of the group allows selective protection based on sterics, but there are also examples of chemoselectivity where carboxylic acids are protected in the presence of primary alcohols, and primary amines are protected in the presence of anilines and alcohols. There’s not much data on what the stabilities of the group in these different environments are - silicon protecting groups on nitrogen are usually pretty flimsy, and not many are useful or widely used. Nor are silyl esters a common sight in total syntheses, but BIBS could be potentially useful for this – apparently these compounds even survive aqueous workups with mild acid or base and can be columned. I suppose, as with all new protecting groups, this information will only come with time.
1. A comparison of the relative acid and base stability of the common silyl groups can be found in a useful (and free!) handout on protecting groups by Andy Myers at Harvard.
2. J. Am. Chem. Soc., 1997, 119, 7897-7898. I love this synthesis, especially the oxidopyrylium-alkene [5 + 2] cycloaddition, a reaction I planned to do as a masters student many years ago, but never got round to. A more recent synthesis of the structurally related daphnanes, also by Wender and using a lot of the same chemistry, can be found in the Nature Chemistry AOPs (DOI:10.1038/nchem.1074). I may yet cover this paper as there’s some fine chemistry.
3. The refluxing stage takes 15 hours. That’s probably not even an overnight job in the Corey lab.