[combatwombat]

I have a way of doing this but I feel like there's got to be a quicker route. Here's how I did it:

1. HCN
2. LiAlH4
3. H2O
4. H2SO4, -H2O
5. H2, Pd

(i.e. cyanohydrin formation, reduction, acid-catalyzed elimination to get rid of OH, then H2/Pd to get rid of double bond)

[gfunk]

Wittig with Ph3P=CH-NH2, then hydrogenate?

[combatwombat]

Money in the bank. Didn't know you could make Wittig reagents out of amines.

[gfunk]

I'm not sure if there'd be some problems with that particular type of phosphonium ylide, with a +R electron-donating R-group on the nucleophilic carbon atom.  I would've checked SciFinder, but I haven't returned one of my library books yet so I couldn't access it :-(.

[Squirmy]

An electron-donating substituent doesn't rule it out as the ylide Ph₃P=CHOMe is a nice way to homologate a ketone or aldehyde (i.e. form a vinyl ether that's hydrolyzed to give an aldehyde with an extra CH₂). However, I believe the methyl ether is necessary for kinetic stability of the phosphonium salt (what you make the ylide from). With a hydrogen on oxygen, it would likely decompose to formaldehyde and triphenylphosphine.

I suspect the NH₂-substituted phosphonium salt would suffer from a similar problem, and you can't have methyls on nitrogen in the product....gfunk: if you find anything, I'm interested. I do miss SciFinder

Re: your original synthesis, I think the LAH reduction might be a problem. Generally, cyanohydrins are not stable to basic conditions. The conjugate base can kick out cyanide ion to give back cyclohexanone. Even if this elimination is reversible, the ketone is more easily reduced than the nitrile, so you'd probably end up with cyclohexyl alcohol.

You could cap the OH of the cyanohydrin with a TMS group (or just use TMSCN from the get go) and then reduce. Deprotection of the TMS could probably just go in with the dehydration step.

Since you're looking for another way: I like your retrosynthetic analysis of RCH₂-NH₂ => RCN.  Is there any other way to incorporate the nitrile other than cyanohydrin formation?

[gfunk]

I figured out a work-around to get SciFinder access on my computer without returning my library book :-).  But then, I have this love/hate relationship with SciFinder.

Anyhow, I couldn't find any reference to a Wittig like Ph3P-CH-NH2, but I think I found Ph3P-CH-NR2 where R is possibly an electron-withdrawing group.  That being said, the primary amine has very similar electronic properties to a hydroxyl group, and I found references to Ph3P-CH-OH (which is also available from Sigma-Aldrich).  As well, the methoxy methyl phosphonium ylide has the same +R electron-donating effects and is also commercially available.

Interesting to note, there were quite a number of references of the reaction forming the cyanohydrin and reduction of the cyano group with LAH to -CH2-NH2 while keeping the hydroxyl group intact.

[Squirmy]

Looked for the phosphonium salt at Aldrich...it's part of their Rare Chemical Library. This disclaimer is a little telling:

Please note that Sigma-Aldrich does not collect analytical data for Rare Chemical Library products. All sales are final

I read that as "it might be crap by the time you get it"   Did any of the references use it in a Wittig reaction?

Can you give an example for the LAH reduction of a cyanohydrin? Looking for a citation or a set of conditions.

[gfunk]

Oh, the Rare Chemical Library .

Anyhow, I looked for the reference I found last night concerning the Ph3P-CH2-OH as a Wittig reagent, but I couldn't find it again!  There were references for it undergoing some sort of reaction (didn't look closely at it, but wasn't as a Wittig)... Last night I found one where it attacked a cyclic ketone forming the enol, then tautomerizing to the aldehyde.  *sigh*

Now that I returned my library books and paid my 50-cent fine, I was able to access the references.  As for the LAH reduction, it turns out that the hydroxyl was protected with TMS.

Here are the references for that:

http://home.cc.umanitoba.ca/~umstella/cyanohydrin.pdf
http://home.cc.umanitoba.ca/~umstella/cnh2.pdf

[cundi]

Reductive amination.
Methyl amine, cyclohexanone and NaBH(AcO)3 or NaBH3CN.
Maybe a acid media is needed.