[curiouscat]

One way to do the following step is a Diels Alder with Acrolein



Are there other options? The motivation was getting rid of Acrolein which is a bit nasty to handle.

[discodermolide]

You could always use an acrylate ester then reduce with DIBAL-H to the aldehyde at -90°C. Or reduce to the alcohol and re-oxidise. More steps = more expensive.

[Dan]

I think you're making a lot of work for yourself if you try constructing this any way other than a DA cycloaddition. You could use some kind of acrylate or acrylonitrile and then reduce to the aldehyde afterwards. These are still toxic reagents, but less volatile than acrolein, so maybe easier to handle, but probably not worth the extra step.

One option might be acrolein diethyl acetal, though it will probably be much less reactive than acrolein.

Another possibility might be a DA 3,4-dihydroxy-1-butene (which is commercially available and non-toxic) followed by periodate cleavage to the target aldehyde. This alkene will be much less reactive though, and will probably require a catalyst which in turn probably requires protection of the diol (Murphy's law). The periodate cleavage also generates formaldehyde (toxic and volatile) anyway.

I would bite the bullet and use acrolein - provided you take appropriate safety precautions, it isn't all that bad on small scale.

[curiouscat]

I would bite the bullet and use acrolein - provided you take appropriate safety precautions, it isn't all that bad on small scale.

Thanks Dan!

I should have mentioned: This is on the small scale right now; but the idea is to scale it up to batches of ~1 ton or so. i.e. commercial processes.

Nevertheless, I might still have to bite the bullet and stay with Acrolein; though I want to make sure there's no safer alternative I forget.


PS. The starting material itself is not constrained so long as it makes economic sense.
 

[curiouscat]

You could always use an acrylate ester then reduce with DIBAL-H to the aldehyde at -90°C. Or reduce to the alcohol and re-oxidise. More steps = more expensive.

Thanks @disco!

[discodermolide]

Don't fancy acrolein on a ton scale, acrylates may be somewhat better, especially if you use an ester that is NOT methyl or ethyl. Other esters then may be more expensive.
I don't know if acrylic acid/sodium acrylate will work, it should do, then you may be able to do the thing in water.

Note: "The WHO suggests a "tolerable oral acrolein intake" of 7.5 mg/day per kilogram of body weight." It is also genotoxic. The maximum allowable concentration in the air is 0.1ppm, 0.25mg/m³. So if you can get it transported to you then it's space suit time. I think that's worse than chloromethylmethyl ether.

[curiouscat]

Don't fancy acrolein on a ton scale,

Me neither.

So if you can get it transported to you then it's space suit time.

Wonder if in-situ or on-site generation of Acrolein might make things any better.

http://www.orgsyn.org/orgsyn/orgsyn/prepContent.asp?prep=cv1p0015

Apparently, all I need is glycerol and KHSO4 etc. I'd rather store large quantities of glycerol than acrolein.

Just brainstorming for now. What do you guys think?

[discodermolide]

I like the comment "After the first violent reaction has abated (about one hour) the baths under B and D are raised to 215–230° ".
Looks like you don't have an addition controlled process here?

[curiouscat]

I like the comment "After the first violent reaction has abated (about one hour) the baths under B and D are raised to 215–230° ".
Looks like you don't have an addition controlled process here?

I have no clue! Although I did read that comment and it was no confidence booster.

It'd the devil n deep-sea choice again, eh?

[discodermolide]

Give it a try with acrylic acid, or the sodium salt. Then you should be able to do a Rosenmund reduction to the aldehyde via the acid chloride.

[curiouscat]

Give it a try with acrylic acid, or the sodium salt. Then you should be able to do a Rosenmund reduction to the aldehyde via the acid chloride.

Thanks again @disco:

[discodermolide]

Yes except the second compound should be an acid chloride.


Then to the aldehyde http://en.wikipedia.org/wiki/Rosenmund_reduction