[zeropoint]

So I have this ridiculous problem which has eaten up my last few days. I have a caffeine like molecule and I need to cleave the two amides on the 6 membered ring to liberate the amine. Conditions that I have found suggested in the literature include:

1) NaBH4 in 6:1 ipro/H20 for 4 hours, then reflux overnight at pH 5 (AcOH)
2) Hydrazine, methyl hydrazine, hydrazine monohydrate, etc. all excess, in MeOH, or in THF, or neat, ON @ RT
3) TFA in both MeOH and neat, ON @ RT
4) LAH in THF, ON @ RT
5) Excess n-BuLi in THF, ON @ RT
6) Excess LiOH with iodine, 4 hours @ RT, then acidify with AcOH
7) 6% HCl/H2O2, in H20, ON @ RT
Excess NBS in THF, ON @ RT
9) Excess NaOH, ON @ RT

As you can see, I have been using some pretty forcing conditions. I get a lot of new spots on TLC (4:3:1 CHCl3/MeOH/NH4OH) but nothing that corresponds with the liberated amine (I have am authentic standard).

Next step is to prep TLC all of this garbage and do the NMR and MS one by one. This could take weeks.

If anyone could suggest a set of conditions to cleave this type of amide, I would be forever grateful. I suppose that this amide is so robust compared to others because to cleave it, you have to disrupt the aromaticity of the entire ring system.

One thing that I thought about doing was using Pd on C and H2 to reduce the ring and then the amide will be easier to cleave... the only problem being that I have a olefin elsewhere in the molecule that I can't sacrifice.

Thanks in advance.

[Doc Oc]

Can you post the structure of your compound?  Without knowing what other sensitive functionalities you're dealing with this is an impossible question to answer.

[zeropoint]

Sure thing.



My compounds are all identical to caffeine (shown) except that instead of a methyl on the diamide nitrogen (top left nitrogen) there are a variety of groups. The most sensitive being an allyilic alcohol. Finding gentler conditions that will spare this functionality is secondary. Right now, I would be very happy with simple cleaving those amides to release the amine.

It occurred to me that one strategy may be to reduce the amides to hemiaminals (say a DIBAL-H reduction at -80C) and then add acid to liberate the amine. Again, the problem being that the ring is aromatic and would prefer to stay that way and so the energetics may not be favorable. I wonder if there is are conditions that can help negate the stability of aromatic systems.

Any help, suggestions, or comments would be hot.

Thanks.

[Doc Oc]

You're trying to break open a purine?  Yikes, good luck with that.

There are multiple problems with this route, the least of which is the relative unreactive nature of an amide.

1) Any reaction you find that can open the amide will not be selective for one amide over the other, so you'd get a mixture of products.  You can try to rationalize that one of the carbonyls is conjugated with the imidazole and the other isn't, but amides overall are so unreactive I don't imagine that would be a significant factor.

2) Anything that can open this ring will potentially also just release your amine from the whole molecule.  Think about it in terms of what you were thinking with the hemiaminal.  Choose one carbonyl, either one.  If you somehow reduce it, it will open into an aldehyde and an amide.  But the reduction can still occur on the other carbonyl after that and if it does, your amine will just pull apart from the molecule altogether.

About the only thing I might be able to suggest is LiAlH₄ reduction.  This would leave you with the piperazine though, not a non-cyclic amine.  I'm not an expert with nucleotide chemistry though so I don't know if this would even work.  I couldn't find a procedure on SciFinder, so I think you'd be on your own.

My personal opinion is that there are probably easier ways to access what you want and you should explore those options.

[zeropoint]

Thanks for your opinion on this.

I may not have been clear. I do want to liberate the amine entirely. The left over remnant of the xanthine ring is of no importance.

The trick will be reducing to the hemiaminal and avoiding the piperazine (as this would trap the amine). Although, with the piperazine, you could then alkylate and make the amine a leaving group... which might be worthwhile some time int he future.

I WISH that I could take another route. There is no chance of that as the xanthine is doing something extremely important in a previous step.

Any more insights would really help.

Thanks a lot =D