[The results of spending years of time, effort and money on this problem. PLEASE DO NOT COPY AND PASTE MY WORK TO OTHER SITES! I will be very disappointed to see it on less scientifically worthy fora.]Here's my reply, two years late. I only just found this.
I believe I know the particular substrate you're interested in performing this workup on. And why.
I have tried running the workup using an AlCl3/DCM system. From what I can tell, it horrifically fails. The result is a filthy dark brown / orange tar like liquid from the DCM phase once the solvent has been evaporated.
Studying other work that uses the same system, some reflux and others can run at room temperature but, universally, the substrates being used lack any other functional groups, leaving the ArylOMe ether as the only target for the lewis acid to work against. They're almost all simple aryls with just the ether group present. Even when the substrates get bigger, they seem to lack other functional groups in all the other work I've seen.
With the substrate I believe a number of you are interested in, there is a double bond hanging off the end of the tail in a primary position, which forms an excellent point for the lewis acid to attack, causing cylic problems and potential dimerization. The lewis acid attacks the ether group by chasing after the spare pair of electrons left orbiting the oxygen; which you see if you look at a molecule as simple as water. The double bond has a similar clustering of negative charge that will also seem appealing to the acid. None of the reagents I've seen so far have been particularly selective when it comes to choosing which to go for under standard reaction conditions.
This particular substrate is very, very picky and troublesome to work with. So picky it's actually used as an example were common workups fail when run against it, even by journal entries.
Heating the substrate, for example, can be enough to cause rearrangement. And other reactions I've seen have had problems with the tail bending all over the place.
When we (I) realized the double bond may be causing problems by supplying a secondary source of electrons, I began thinking of easy and productive ways to lower it's potential to do so. The first was to use the iso-substrate, featuring a double bond in the secondary position. Another, and probably better, method was to convert the double bond to a ketone, using the electronegativity of the oxygen to secure the charge in that region.
On attempting to transform the substrate to the ketone using the wacker, I discovered that this is yet another workup the substrate isn't at all interested in. The catalytic oxidation doesn't work because the catalyst / substrate complex that needs to form for the controlled oxidation, doesn't. It ends up a mess and the substrate is oxidized to a tar in an uncontrolled manner. There is actually a patent out there suggesting that high yields can be had from the wacker when run against this substrate. I tried it multiple times following their workup to the letter. Every time, it failed. I also tried a ketone reaction vessel like approach, as well a number of other oxidizers.
Afterwards, I was made aware of a journal entry detailing the failure of the substrate in the wacker.
I've lost count of how many times my glassware ended up caked in tar that I could barely get off with boiling saturated KOH; aggressive enough it would dissolve my cleaning brushes and yet still be unable to budge the tar without hours of scrubbing, shaking and rewashing each time the wacker failed. I lost some very expensive glass and equipment in the process and had a microwave reactor's glass interior explode at one point, making a bang loud enough to temporarily deafen me in one ear and sending shards of glass flying everywhere. I wasn't impressed, and was very angry given how little help my colleges were giving me; which has been essentially zero since the more important resources, that were previously spoon feeding them the work, went missing. As a result, I left the field of work alone for a long period and all be gave up.
I think this whole area of wacker oxidation of this particular substrate is, unfortunately, a none starter.
I believe it is due primarily to there being an OH group present on the aryl side of the molecule, which is interfering with the catalyst complex as it tries to form. So when you try to work on one end for the demethylation, the double bond gets in the way. And when you try to work on the double bond, the aryl side gets in the way. Boy oh boy...
However, provided the demethylation can be made to function and the aryl end of the molecule stabilized, the wacker will again start functioning; as there is a detailed history of it performing on the derivatives that feature higher stability at the aryl side.
An alternative, to stabilize the molecule prior to a demethylation, may be to go from the iso form to the ketone via a peracid oxidation. This would bypass the problem caused by the standard catalyst / substrate complex failing to form and, hopefully, not but caused to fail it's self due to the functional groups on the aryl side of the molecule. Lacking any iso substrate, I'm unable to test this at the moment. I'm also unable to test whether or not the ketone will even function in a standard AlCl3 demethylation since I was unable to get the substrate through the wacker. But this forms another point where others could optimize the work if they are successful with it. For example, should I stablize and then demethylate or demethylate and then stablize; if it can be done either way round, great, but if that's true, it's also likely doing one before the other will provide higher total yields and perhaps a more simplified workup.
A BBr3 demethylation has been suggested to me as, apparently, it is less prone to attacking double bonds, I am unsure precisely why that may be at present. I am willing to try the BBr3 method, but I don't have any around and I'd have to make some up, which will be a major pain if it also fails.
Worthy of a note is that, reading another chemistry message board, I discovered a thread containing complaints from other chemists about what the messiest workups were they'd run. They mentioned BBr3 demethylations.
What was interesting is that they also mentioned how temperature affects the process and that the more prone the substrate is to failure, the cooler you should run the reaction.
It may therefore be possible to make use of a standard AlCl3 / DCM system provided the temperature is dropped down low. I may give this an attempt, as I have some of the materials left from a previous attempt at room temperature.
Of interest is that this advice is in direct contrast to what a lot of people would suggest, that the reaction needs refluxing to run.
If AlCl3 can be made to perform on the substrate, this is great news, as it's also extremely easy to make in large quantities, with demethylations typically requiring a molar excess of the lewis acid from what I can tell.
I've seen all the other methods, like pyridine. I've even spoken to someone who's tried hard acid / soft base attacks with AlCl3 and Thiourea. I have some thiourea handy, but I don't think the work up is failing due to a lack of capacity from the AlCl3 alone, I think it's far more likely a temperature related issue.
If it works, it may also need some tedious experiments with temperature to find an optimum, as the reaction may not have the activation energy needed at the lower temperatures and then, as they climb, will start attacking the double bond again. I may be able to get hold of some dry ice, but I'm running desperately low on reagents as well and I need to hydrolyze the reaction to check for products, ruining the lewis acid, so I'll have to plan out the sequence of tests careful to try and cover both -78 and -40C. 0C may also work. But I am relatively sure standard room temperature is too much.
My attempts were using PTC.
It may also be worth switching the solvent from DCM to a true none polar. In some of the work I've seen, using something like cyclohexane should result in the demethylated producted immediately becoming visible as a solid ready for filtration after the hydrolysis. Using DCM causes it to go into solution in that phase. Although I'm not sure how swapping from a polar aprotic to a true none polar will affect the overall efficiency of the reaction. I also don't have any cyclohexane to hand.
I can also tell you the clean up routine for doing it with DCM, as per the expensive journal entries I've bought. It's very simple and has nothing to do with A/B extraction. You simply hydrolyze the AlCl3 / DCM phase. The two will split beautifully, as DCM / Water washes always do. Then you can either acidify the aqueous phase to 5% with HCl or, and probably better, get rid of it and rewash it with some clean water. This washes byproduct into the aqueous layer, not the product it's self, which remains in the polar aprotic solvent. Dry the DCM with however much Mag Sulfate you need to get rid of the remaining drops of water that are left floating around, boil off the DCM. Done.
Given how prone the substrate involved is to problems related to heat, and the product is even more so I believe, it may be best to let the DCM evaporate at room temperature (with a fan over an evaporating dish) or boil it off under vacuum, without any heat. Since DCM will strip off at such low temperatures, and boils so fiercely under vacuum, that should be easy.
I've heard suggestions this kind of thing should be done, the entire workup and filtration, under an inert atmosphere. Personally, I'm not so sure how important that is provided you're not trying to squeeze the last 1% of conversion out of it. Your lewis acid MUST, be it's definition, be the anhydrous form. If your substrate or DCM are wet, it'll hydrolyze some of the acid into the salt. But again, I'm not sure how important it is to have everything perfectly dry provided you're only doing it for a dirty yield and not precision.
It may even be possible to run the workup on a dirty substrate direct from an organic source, reducing the number of vacuum distillations - since you'd still probably want to vacuum distill the product even if running from a pure form of the substrate.
If the workup fails to operate at lower temperatures, but not tar is produced, it may be that adding a soft base, like Thriourea, is what's needed to get it operating at the temperatures that won't simultaneously ruin the substrate.
I hope this helps some of you.
I think this area of chemistry has lost a lot of what it had since the collapse of the most important websites discussing it a while ago. I've yet to meet many people who wish to restore the same mentality or who are interested in doing any of this themselves. But this time, please lets not be publishing anything that could be so easily abused or writing any books on the subject that could so easily lead to the same thing happening, again.
Obtaining the results I have has used up a lot of my money, time and effort, and so I am reluctant to hand them out for other people to profit from. I've been working at this problems for years and have ruined a lot of my life by paying so much attention to it and not enough to my friends. But this isn't getting me / us very far. So I ask that you please remember this fact if you're in a similar conundrum when considering how much of your work you wish to discuss with me. I have tried hard for myself and you guys, please don't steal that from me; I mean, take it and give nothing back. I would also appreciate it if you can credit me for the work when quoting it; a minute price to pay when compared to what it's cost me to provide it to you.
If you're still interested, please reply just to let me know; I'd like to know someone is reading. I'd also love to hear any newer ideas or stories about work with this idea.
All my love and a big smile,
Peach
ChemBuddy | 
Chem Reddit | 
Topic: Demethylation by BBr3 (Read 44444 times)