Sciencemadness Discussion Board - LiOCH3 to replace NaOCH3?

LiOCH3 to replace NaOCH3?

Intergalactic_Captain - 19-6-2007 at 11:33

Question is pretty straightforward. Desired end product is syringaldehyde from 5-bromovanillin. All my refs use a solution of NaOCH3 in methanol, or a 2-step to the phenol followed by methylation.

Aim is to avoid methylating agents and directly methoxylate the aryl halide. Due to lack of resources to build a castner cell, I would like to know if the lithium alkoxide is a viable alternative.

Also, for the sake of asking the question, does anyone know how much lithium is in a AA size Energizer E2 battery? The only thing I've found on lithium reclamation is the rhodium doc. on erowid and no yeild is listed.

Nicodem - 19-6-2007 at 12:00

You can use NaOH in methanol. The yields will be a little lower though.

Intergalactic_Captain - 19-6-2007 at 12:09

Interesting... Would you happen to have any refs on this?

smuv - 19-6-2007 at 21:55

If I recall correctly...in a old text book I once read (and it makes sense), that if KOH or NaOH is added to dry ethanol and to this solution acetone is added, the alchoxide will precipitate out.

EDIT: I of course assume this could be applied to methanol as well.

[Edited on 19-6-2007 by smuv]

Intergalactic_Captain - 20-6-2007 at 06:10

The problem with alkoxide formation is that it is basically an equillibrium reaction with water forming mole for mole. The water formed "poisons" further alkoxide formation.

The reason it works with ethanol is that they assume lab-grade or anhydrous to start. Ethanol forms an azeotrope with water at ~95% EtOH which boils slightly lower than pure ethanol. As such, water formed in the reaction is removed by boiling it out as it is formed.

No such azeotrope exists for methanol. A tiny amount of methoxide is formed, though, something like .01%. The water has no way to get out of the reaction mixture, and thus the equillibrium cannot be shifted any further to the right.

This is why I'm wondering about the method Nicodem posted - Are you referring to forming the methoxide, or reacting the bromovanillin in a NaOH/MeOH solution? Either way, reaction conditions, yeilds, or any other notes would be appreciated.

Nicodem - 20-6-2007 at 10:33

Quote:

Originally posted by Intergalactic_Captain
The problem with alkoxide formation is that it is basically an equillibrium reaction with water forming mole for mole. The water formed "poisons" further alkoxide formation.
...
A tiny amount of methoxide is formed, though, something like .01%. The water has no way to get out of the reaction mixture, and thus the equillibrium cannot be shifted any further to the right.

Instead of guessing and even erroneously so, you could have checked the pKa values of MeOH and H2O and calculate the ratio (the pKa are 15.5 and 15.7 correspondingly in water as solvent).

Thus the equilibrium constant for:
MeOH + OH- <=> H2O + MeO-
is K = 10(15.7-15.5) = 1.6 in water as solvent.

In your case where the solvent is MeOH, the K should be a bit lower. I don't have the pKa data for MeOH and H2O in methanol, but I would estimate the K is between 0.5 and 1.

In the solution of NaOH in MeOH the [H2O] = [MeO-] due to the reaction stoichiometry. If you put this in the equilibrium equation you get:

([H2O]*[MeO-]) / ([MeOH]*[OH-]) = K

[MeO-]2 / ([MeOH]*[OH-]) = K

[MeO-]2 / [OH-] = K*[MeOH]

Normalizing by [OH-] = 1 you get the ratio for
[MeO-] : [OH-] = sqr(K*[MeOH]) : 1

For low enough concentrations of NaOH the [MeOH] is aproximately the concentration of pure methanol which is 24.7 mol/l.
Let's take an average estimate for K=0.75. Therefore the ratio
[MeO-] : [OH-] = 4.3 : 1

Even at the K = 0.5 you still get the ratio of 3.5 : 1 in favor of the methoxide, while at the K(aq.) = 1.6 you get 6.3 : 1. So even at the worse estimation of K the methoxide anion prevails over the hydroxide in the solutions of NaOH in methanol.

That is not particularly surprising since in the reactions where the presence of water has no bad influence on the reaction (like yours) one can use solutions of hydroxides in methanol if the methoxide is the desired nucleophile for the reaction. For example, the solvolysis of benzyl chloride in NaOH/MeOH gives almost exclusively the benzyl methyl ether (PhCH2OMe) and much less benzyl alcohol. Of course in a lab one would use NaOMe since it is almost as common as NaOH.

PS: I would appreciate if you would do the calculations by yourself next time as I really hate doing them.

Intergalactic_Captain - 20-6-2007 at 11:48

Wow...Thanks. One thing, though, the portion you quoted from my last post was focused on isolation of the methoxide, not merely its formation. As far as I am aware, it is damn near impossible to mix up some NaOH, MeOH, and evaporate to get any substantial amount of NaOMe. On the other hand, according to your numbers, there should be no problem with using this solution as a reaction medium.

And I did kind of pull the .01% out of my ass - Well, not out of my ass, rather it was from a Hive Archive post. Equillibrium math is not one of my strong suits, but I could get better with a little practice - Only problem is finding pKa data. I don't have a CRC handbook and my Merck index is horrendously outdated. Is there any list of sorts of common pKa and pKb values out there somewhere?

More to the point of the thread, is there any reason why LiOMe would not work? The one thread on the hive regarding it brought up the point that its max concentration in methanol might be too low to facilitate the reaction, though nobody actually tested it out. I suppose methanolic lye is about as easy as it gets, but if a pure methoxide solution gives better yeilds it would make for an easier cleanup.

EDIT - Found what looks like the best method yet - Quantitative preparation of metal alkoxides from alcoholic hydroxide solutions via dehydration by zeolites

[Edited on 6-20-07 by Intergalactic_Captain]

Attachment: PREPARATION_OF_METAL_ALKOXIDES.pdf (82kB)
This file has been downloaded 832 times

smuv - 20-6-2007 at 21:33

I gave you a to isolate pure alkoxide, by precipitating it with dry acetone.

Nicodem - 21-6-2007 at 11:06

Smuv, the patent US1978647 you are (unsuccessfully) referring to has been already discussed in the alkoxide thread and I'm pretty sure Intergalactic is familiar with it or else he should UTFSE anyway.

Quote:

More to the point of the thread, is there any reason why LiOMe would not work? The one thread on the hive regarding it brought up the point that its max concentration in methanol might be too low to facilitate the reaction, though nobody actually tested it out. I suppose methanolic lye is about as easy as it gets, but if a pure methoxide solution gives better yeilds it would make for an easier cleanup.

I don't know of the exact solubility of LiOMe in MeOH, but I would assume this should be no problem given the solubility of almost anything tends to increase with temperature and since your reaction takes place at 125-150°C…
By the way, what kind of autoclave are you going to use? I hope you realize that methanol at that temperature has the partial pressure of nearly 20 bar and I would hate to hear a forum member got injured due to a faulty homemade autoclave. Alternatively use the method posted here by Ullmann where he uses DMF as cosolvent, which besides increasing the reaction rate also reduces the partial pressure in the vessel.
Also what kind of Cu ligand do you intend to use? You do realize that unless you perform the reaction in DMF or NMP you need some copper ligand or else the conversion will suck?

PS: Using NaOH/MeOH would be lower yielding (if at all), but you would at least do some research useful for others. Besides you will have to recrystallize the product anyway and when recrystallizing from water/ethanol the much more water soluble 5-hydroxyvanillin will remain in the filtrate. I would say that 5-bromovanillin, vanillin and other side products are more of a trouble to separate completely.

[Edited on 21/6/2007 by Nicodem]

Intergalactic_Captain - 21-6-2007 at 12:10

This may be a stupid question, but why would I need an autoclave? Beilstein hits only showed 125-150C to be necessary when using NaOH/MeOH. The reactions done using NaOMe/MeOH were all stated to be done at reflux, occasionally with a DMF cosolvent. Would LiOMe require pressure-vessel conditions to react?

I've got some 3A mol sieves coming in the mail soon and am going to try out the NaOH/MeOH dehydration outlined in the patent I attached...I've run this through my mind many times in the past, but it was only recently that I found the patent for validation of the procedure. Should this fail, I'll give the LiOMe/MeOH a shot, and failing that the NaOH/MeOH. I've no real autoclave, only a 3/4" wall-thickness aluminum pressure vessel - I'd give it a go first if it was stainless, but since it's aluminum anything with NaOH is a definite NO. If I can find a suitable way to plate it with copper it may be a feasible vessel, but for now its out of commission.

EDIT - Forgot about the Cu...I'm looking at this
[Edited on 6-21-07 by Intergalactic_Captain] method, mainly;

Quote:

Another example, of industrial interest, is the transformation of 5-bromovanillin into syringaldehyde: substrate (5 mmol) is refluxed with EtOAc (3 mmol) and CuBr (1 mmol) in 5 M MeONa/MeOH (10 ml) for 14 h; classical work-up leads to pure syringaldehyde (95%). When starting from more soluble 5-bromovanillin dimethyl acetal, reaction is achieved within two hours (yield 98%).

Not sure how to aquire the CuBr yet, but it looks like a facile enough procedure. If I can find a cheap civilian source I'll just buy it, but otherwise it'll have to be homemade.

[Edited on 6-21-07 by Intergalactic_Captain]

Nicodem - 21-6-2007 at 12:39

Edit: Ah, you talk about that Tetrahedron Lett. paper with EtOAc as proligand. Well, that's about the only paper with reaction conditions mild as that. All the rest I checked were >110°C. (BTW, you can not use NaOH for that one since EtOAc hydrolyses in the presence of water.)

Quote:

Beilstein hits only showed 125-150C to be necessary when using NaOH/MeOH.

If you found a reference where NaOH/MeOH is used in the preparation of syringaldehyde then why you had so many doubts? Please provide the reference.

[Edited on 21/6/2007 by Nicodem]

Intergalactic_Captain - 21-6-2007 at 12:49

You were typing while I was editing...Otherwise you'd have seen the one I just posted. Also, my doubts were mainly to whether my ability to perform the NaOH/MeOH reaction with my current resources would be successful. Secondarily, they were based on the fact that water would not be removed from the mixture, keeping the methoxide formation equillibrium far to the left - I may be wrong in my line of thought, but would much rather try a method with at least some academic backing...I've no analytical equipment, not even an mp tester, to verify what's in the jumble of reaction products.

I ran another beilstein search, and found what I was looking for - Yes, they all run at 110degC+...The ref I was thinking of was one regarding 2-nitro-p-dichlorobenzene, which I believe I posted in a thread dealing with that compound.

EDIT - Wow...We're both typing/reading at the exact same time...These damn edits are getting annoying...

Regarding NaOH in the reaction, the current plan is to use NaOMe - Check out the patent I posted regarding methoxide formation w/ zeolites. There isn't too much data regarding yeilds or purity, but in theory using an excess of the amount of zeolite required to remove the water formed by the reaction should push the reaction equillibrium to completion, forming a NaOMe/MeOH mixture that is practically free of NaOH. Of course, this solution would be formed beforehand and then used for the methoxylation reaction.

[Edited on 6-21-07 by Intergalactic_Captain]

Nicodem - 21-6-2007 at 12:59

The reaction of methoxide with 2-nitro-p-dichlorobenzene to yield 2-nitro-4-chloroanisole goes by an addition/elimination mechanism known as nucleophilic aromatic substitution and is absolutely irrelevant to the reaction you want to acomplish!
You should be looking for Ullmann aka Goldberg aromatic substitutions which proceed by an SRN1 mechanism.
It would sure help if you would at least know what you are doing.

Intergalactic_Captain - 21-6-2007 at 13:30

Yeah, a little more knowledge of what I'm doing wouldn't hurt...Only problem is that I'm usually thinking of so many different things at a time that confusion is pretty much a given...I try not to slip up too much, but this thread is a prime example of what happens when I've got 4 different projects in mind with completely different pathways/products/mechanisms.

Ullmann - 24-6-2007 at 12:48

Check this link:

Synthesis 308 1983

and that thread:

Ullmann m-escaline

The methoxyde solution used in the above thread was done from KOH/MeOH + 3A sieves..., first dissolve the KOH in boiling anhydrous methanol to make at least a 3M solution (cautiously, it is quite basic and dissolve slowly), let it cool, decant from the residue that may be present and poor the solution in a capped container. Add the sieves (100g / mole water), it heats up, let the solution sit at least for a night before using. Keep tightly closed. Yields are good by using this solution for ullmannisation.

Personal attempts to use the undried methoxyde failed. Maybe it was a catalyst issue i do not remember.

Personal attemps by using the EtOAc protocol failed.

DMF as a cosolvent is best used, NMP can be substituted.

An autoclave is by no way obligatory, just reflux for a few hours and use a guard tube it will be ok.

A copper one salt such as CuBr is best use, nonetheless you can substitute it by a copper two salt and ascorbic acid to reduce it.

A good reaction is brown at the end, a failed one is blue. Bromovanillin precipitate readily from the failed reaction by addition of water and acidification, syringaldehyde do not do this if it is formed and that means the reaction is succesfull.

The product typically contain from 5 to 10% vanillin as a dehalogenated product, it is the only contaminant and bromovanillin is nowhere to be found for a succesfull reaction.

Recrystallisation from toluene can be used to increase purity but if you do not care the crude product can be used for the next step, then the 4-alkoxyaldehyde or the nitrostyrene (at least!) has to be recrystallised.

good luck

Nicodem - 24-6-2007 at 13:35

Quote:

Originally posted by Ullmann
Personal attempts to use the undried methoxyde failed. Maybe it was a catalyst issue i do not remember.

You were using it with DMF, right? In such case it would be unlikely to work anyway since the DMF hydrolysis rate is probably of similar speed as the Ullmann reaction. I still think it would be nice to try with only MeOH as solvent provided the phenolate is preformed dry by neutralizing in i-PrOH with KOH/NaOH and rotavaping it well. Of course, one of those patent autoclave procedures would need to be used which is a quite unpractical for the avarage amateur. But then again, molecular sieves become more and more accessible...

Quote:

An autoclave is by no way obligatory, just reflux for a few hours and use a guard tube it will be ok.

That is only the case if you use DMF or NMP as cosolvent in which case the reflux temperature can well reach above 100°C, depending on how much MeOH is added and how hard you heat to keep the MeOH vapors above the solution. But with MeOH as solvent like Intergalactic wants to play it, all the references I saw state a temperature in average of 125°C. Perhaps it might be possible to lower the reaction temperature to the MeOH reflux by adding a super efficient Cu ligand. Perhaps it would work with proline, but I kind of doubt it as it would really have to do a miracle since by a rule of tumb going from ~120°C to ~70°C requires a 105-times increase in the reaction rate (this and your report is also a good reason for having serious doubts in veracity of the paper where AcOEt is used as a ligand precursor).

Quote:

A copper one salt such as CuBr is best use, nonetheless you can substitute it by a copper two salt and ascorbic acid to reduce it.

In my experience with Ullmann reactions, 5-15 mol% Cu powder and CuCl or CuI (1:1) works good, though I never compared it to a reaction without Cu powder. The idea is to prevent the oxidation of Cu(I) to Cu(II) thus increasing the available Cu(I) throughout the reaction time. Cu(II) salts could perhaps be used by in situ reducing them to Cu(I) by the addition of Cu powder (instead of using ascorbic acid, I mean).

By the way, thanks a lot for sharing!

[Edited on 24/6/2007 by Nicodem]

Ullmann - 24-6-2007 at 14:33

Quote:

It was a long time ago, but i think i have tried it both without DMF then with DMF (latter), none of those reaction did work. BTW none of the ullmann reaction i did without DMF or NMP worked, not even if the volume of DMF to the 3M methoxyde solution is like 1:10, and by using an autoclave at 130°C it never worked. The idea to put less DMF (or no) and incresing temperature came from a french patent from roche iirc, they claimed it worked in 4 hour at 130°C under autogeneous pressure but i never succeeded to reproduce this. When it first worked was when i increased the volume of DMF to say 1:1 with regards to the 3M sodium methoxyde solution.

I had tested a few copper two salt, especially the sulfate (dry) as it is easily accessible and the basic carbonate as in the patent they said it is ok to use it. Never it worked for me with those salt (unless ascorbate was put in the medium). I never tried to put metallic copper as a reducing agent though.
Hence the reaction that let me thought undried methoxyde do not work may have been one where such a copper carbonate salt was used, that is what i have forgot. One day i could regive it a try to see, especially since there is a tetrahedron article where they purposely put water in their dmf for the ullmann methoxylation, they said it gives better yields when it is abit wet. That is why i am not sure what i actually did that time. But what is certain is by using sieves dryed potash in meoh the reaction work great. And i am sure the reaction is not -that- sensitive to the presence of water, but may fail if it is too much wet.

The reaction ask for a 3 molar solution of methoxyde to be used, thats because the excess methoxyde apparently reduce the copper two to copper one back, this do not occur at say one molar sodium methoxyde solution.

By using the reflux method, it is best to slowly distill the methanol out of the reaction, as it increase the concentration of methoxyde and the rate.

I do not think proline will work good as it did not well in some article about copper catalyzed azidation of bromo-aryl, a good ligand to try would be sym-dimethylethylene diamine as it gave excellent result in this reaction (which is quite similar to the ullmann). But DMF or NMP are quite accessible solvent and works great.

BTW if you know a welder it is quite easy to get a simple bomb as an autoclave. The autogeneous pressure of MeOH at 120-130°C was between 5 to 6 bars, no more. As long as the screws are strong it is ok. Mine has got only a cap that i can screw, with abit of teflon tape there are no loss of solvent. It is a useful reaction vial for ullmann's.

[Edited on by Ullmann]