[Milpool]

Hi,

I'm currently trying to convert cyclohexanone to 2-chloro-1-formyl-3-hydroxymethylenecyclohexene. It's essentially a double Vilsmeier reaction, and I'm refluxing a large excess of DMF and POCl3 with cyclohexanone in DCM for a few hours. All the protocols I've seen say that it can be precipitated in high yields by pouring the reaction mixture into ice or cold water, but I've found that the DCM just seems to form a dark oily layer at the bottom while the water turns very orange.

I tried cooling it and boiling off some of the DCM to get it to crash out, but I only managed to get a tiny amount of olive green precipitate that quickly turned brown (product is apparently yellow). I also tried extracting the aqueous layer with DCM, evaporating it, and precipitating from the resulting oil with pentane (one of the sources suggested this), but it did nothing.

I finally removed the pentane and mixed the oil and water again, and tried adding some NaOH to increase the pH, which was around 0, likely from the POCl3. This immediately caused a lot of light yellow powder to crash out, despite not changing the pH much. I kept adding more NaOH and filtering off the precipitate, and I eventually started getting the green stuff again.

I took a quick LC-MS of both the green and yellow precipitates, and both seemed to be reasonably clean, but they both had m/z values of 187 (the desired product should be around 173).

I haven't had a chance to get NMRs yet, but I was wondering if anyone was familiar with what might be going on. I found dozens of papers claiming to have precipitated the product without any issues, so it seems a little odd.

This is a link to the primary source btw, every other paper basically leads back to this one, and they all follow it nearly exactly.

https://dx.doi.org/10.1021/jo00425a027

(mod edit, link)

[Mitch]

I'm not familiar with this reaction. But the solvents you used for the first step were completely dry and distilled, right?

[Milpool]

I didn't actually, although that's a good point, I guess I've been doing a lot of non water-sensitive chemistry lately.

It definitely seems like the reaction was going forward, based on the colour change, and none of the sources mentioned anhydrous solvents, but I can see how water might interfere.

I'll give it a try.

[phth]

Mitch is right. Distill everything or else reproducibility is a problem. Anhydrous solvents change the rate of the reaction and probably the outcome of the reaction when conducting VH formylations e.g. hydrogen chloride.

[wildfyr]

In a fascinating coincidence I have recently put a whole bunch of effort into studying this reaction myself, and have seen this exact paper! Did you check the Zemlick and Arnold paper they cite as inspiration?

I found it here
http://cccc.uochb.cas.cz/26/11/2852/ (its in english, thank god)

Reynolds say they are repeating the Zemlick paper, but it seems Zemlick used phosgene instead of POCl3 (!). I didn't find much useful stuff there, but you may catch something I didn't.

Anyways, removing DCM will not cause it to crash out, because DMF is what is left, and it's a great solvent. I would suggest removing DCM and adding water, which I believe is the intended protocol, perhaps stated unclearly. This will crash out any organics you care about and leave any salts or charged intermediates in the aqueous phase.

Please let me know how this work moves forward, I have been contemplating using this reaction but have yet to buy reagents and take the dive.

P.S. Does anyone else find it curious that the cyclohexanone gives the bis aldehyde, but the cyclopentanone gives the bis-imine in that paper? I wish they had explained exactly what they did, as it seems some extra reaction was done to get the bisimine. I'm particularly interested in the cyclopentanone route.

P.P.S. Hi Mitch, nice to finally meet you.

P.P.P.S Posts went up while I'm typing this. Usually I would happily blame water, but frankly I don't think it's the culprit in this reaction unless the DMF is totally soaked.

[Milpool]

I did see the original Zemlick paper actually, but I didn't notice anything super useful either. 

Not sure if I'm misreading your suggestion, but wouldn't removing the DCM prior to adding water still leave the DMF, which could keep the product dissolved? Or would removing the DCM before the water step this at all?

In retrospect, the DMF I used was quite old and stored in a regular bottle, so it's entirely possible that it is soaked.

[crawlingmcedge]

How much water did you throw in? DMF/DMSO are painful to remove. You really do need a significant excess to get rid of it.

Much worse when you need to remove them from water soluble compounds as well.

[wildfyr]

You need to add a whole lot of water. 20x the amount of DMF. This will cause the DMF soluble organics to precipitate. You then wash the solids on the filter several more times with water to remove residual DMF.

I highly suggest you run this in the most simple way possible (don't bother with ultra-dry conditions or new solvent yet. Maybe use some sieves to dry your DMF overnight), but at the end, rotovap off the DCM, then add a whole bunch of water. I'm virtually sure this is what the intended procedure is. Its quite common for using DMF, and is used at several points in this cyanine dye synthesis pathway.

[opsomath]

It's silly to add a DCM-containing mixture to water and expect anything other than an oil. DCM's completely immiscible and heavier than water. So either (1) rotovap for a long time at a high heat and pray, *then* add your crude stuff in DMF to crushed ice to precipitate, or just do a normal water/DCM extraction in a sep funnel, washing several times with water to remove the DMF.

[Milpool]

So I took wildfyr's advice and everything worked out great. I used some fresh, dry DMF (not sure it made a big difference, but just wanted to be safe), and followed the same procedure, except this time I evaporated the DCM before adding the remaining mixture to a lot of ice. Since there was no DCM left, there was no oil layer and it formed a homogeneous orange mixture which formed a yellow precipitate once I stirred. The product was nice and clean and the yield was reasonably close to the literature.

Also, I agree that it's obvious that DCM will form a separate layer when added to water, that's why I was confused upon initially reading the procedure. I still find it a bit odd that literally none of the papers I saw thought to mention this step, but I'll be more mindful of this sort of thing in the future.

Side note: The DCM came off quickly on the rotovap and didn't need any heating above the usual 40 degrees.

[Mitch]

That's great news, I'm glad it worked for you.