Author Topic: Demethylation of quaternary ammonium salts  (Read 131 times)

atara

  • Dominant Queen
  • ****
  • Posts: 256
Demethylation of quaternary ammonium salts
« on: July 11, 2012, 05:34:15 AM »
I figure this paper will help lots of people. Reagents are acetate and a solvent. Couldn't get much easier.

Of particular interest: LSD + allyl iodide --> AL-LAD methiodide --[demethylation]--> AL-LAD... of course, you could also demethylate N,N,N-trimethyltryptammonium to DMT.

Enkidu

  • Global Moderator
  • Foundress Queen
  • *****
  • Posts: 677
Re: Demethylation of quaternary ammonium salts
« Reply #1 on: July 11, 2012, 02:44:59 PM »
The only trouble with this procedure is swapping salts, e.g., I- to AcO-, unless there's an easier method than running a column.  Columns are only easy if the material to be separated is chromatic. I never pursued the swap further, and I'm too lazy to look it up right now.

From what I've found, there are three other facile routes: one involving ethanolamine, another thiourea, and the last sodium thiophenol.

you could also demethylate N,N,N-trimethyltryptammonium to DMT.

One of the main uses around here IMO. Aminoethanol has been reported as a successful reagent for demethylation to DMT.

Organic Syntheses, Coll. Vol. 5, p.1018 (1973); Vol. 49, p.107 (1969). (the attachment called ammonium demethylation ethanolamine)


Murphy was kind enough to translate the Ber. papers for me sometime last year.

Quote from: Murphy, by PM
OK, for both the kinetic measurements (…not interesting for you, I guess) as well as for the preparative application of the dequaternization you’ll need an auxiliary base with a boiling point preferentially above 170 °C and with good dissolving power for quaternary ammonium salts. Both requirements are fulfilled by the three possible ethanolamines. Wiki states as boiling points 172 °C, 269 °C and 360 °C for mono-, di- and triethanolamine, respectively. In particular monoethanolamine possesses the desired dissolving properties. As kinetic experiments demonstrated is this compound also the preferred choice out of the three mentioned reagents for dequaternizations; triethanolamine, for example, has a 50 times lower activity in this respect.

Have a look at Table 2 of the first Chemische Berichte-paper (bottom of p.397): Unsaturated alkyl chains will preferentially leave the quaternary center. Hence, when trying to cleave ?3-butenyl-trimethylammonium acetate you won’t get the ester but exclusively butadiene. I think this detail should be considered in the course of own experiments, as it allows selective cleavage of certain alkyl chains. There are basically 2 possibilities how a sidechain will cleave from the quaternary center:
a) Either as alkene (if the alkyl chain allows this at all), which happens preferentially in presence of strong basic anions like hydroxy or alkoxy,
b) Or the residue to be cleaved joins the anion of the reaction, ie. the nitrogen in ethanolamine (transalkylation).

Now have a look at Table 4: 5-10 g quaternary salt were distilled from a 10-15fold excess of ethanolamine at 154 °C; the amount of cleaved residue was determined by quaternizing again with methyl iodide, which led to precipitation, rendering the isolation very easy. The products are subsequently weighed out. Table 4 shows clearly that when quaternary compounds are employed, in which at least one chain is unsaturated (eg. the first 3 entries from that table), it will be the unsaturated chain that leaves the molecule first; in case of 3-butenyl-trimethylammonium you’ll get almost exclusively trimethylamine!

In general, the rule of thumb goes as follows:
If in case of an "alkyl"-trimethylammonium salt the "alkyl" is an ethyl or even something bigger, you will cleave practically only one of the remaining methyl-groups. This behaviour can be exploited in a preparative manner.

Due to the low basicity of aromatic amines one can cleave off aliphatic chains from the corresponding quaternary salts with particular ease. From trimethylanilinium iodide in boiling ethanolamine you’ll get already after 5 min practically quantitative yields of dimethylaniline.


Experimental Section
The employed amines were of technical grade and, therefore, distilled in vacuum prior to use. The quaternary salts of the type [Me3N-R]+X- were prepared from trimethylamine and alkylhalides (see eg. JACS 1943, 65, p.692; JACS 1952, 74, p.509), purified to the point that literature melting points were met (recrystallisation from EtOAc/MeOH or absolute EtOH) and the halogen content determined by titration. Solely [Me3N-CH2-CH2-CH=CH2]+I- was prepared from dimethyl-?3-butenylamine and methyl iodide. The chlorides are obtained from the corresponding iodides by evaporation of a solution in methanolic hydrochloric acid (see JACS 1952, 74, p.5231). Thus, dimethyl-diethyl-ammonium iodide was prepared from methyl-diethylamine and methyl iodide in 94% yield; methyl-triethyl-ammonium iodide from triethylamine and methyl iodide accordingly with 94% yield.

Preparative degradation and separation of the tertiary amines
The preparative degradation is performed in a round-bottom flask with an attached high distillation bridge, which leads directly to a well cooled receiver. Weighed samples of about 5-10 g salt in an 10-15fold excess ethanolamine are heated for 1.5-6 h under intense boiling. With concomitant reflux of ethanolamine the tertiary amine products of the cleavage distil over into the well-chilled receiver during the course of the reaction. (Note: The authors recommend chilling the receiver with MeOH/CO2. In other words: Ensure that the receiver is kept really cold!)
A short distillation bridge is then attached, with the receiving end reaching below (!) the surface of the solution in the receiving flask, which gets filled with a methanolic solution of excess methyl iodide. Distillation is started by gentle heating, upon which triethylamine and other volatiles are led directly into the reaction solution in the receiver. A water bath can be employed as soon as the more volatile components are distilled off, so that higher boiling amines are carried over, too (optionally, heat with an oil bath up to 120 °C). The exothermic reaction requires that the receiver is chilled thoroughly as long as amines are still introduced into the alkylating mixture. When the distillation has finished, the distillation bridge can be rinsed once with MeOH and the receiver is then closed. Let rest overnight. Already during the distillation, latest after letting the solution rest for some time, the quaternized salts of your products will precipitate.

Separation of the salt mixture
(Note: See also Liebigs Annalen der Chemie 1911, 382, p.149; unfortunately in German, too, I guess)
The reaction mixture is evaporated to dryness with care (CAUTION! Methyliodide is fucking volatile, toxic and carcinogenic!), in the final phase employing an infrared-heater. The salt mixture, dried at 90 °C in an oven and in a desiccator, is weighed out. Dissolve in 40 mL anhydrous EtOH with heating, filter off insolubles through a pre-heated frit and wash with 2x5 ml hot EtOH. The solvent is evaporated off and the resulting salt dried like before. Because 1.0 g tetramethylammonium iodide dissolve in 1060 g hot absolute EtOH, one needs to adjust the amount of salt when using a different volumes of EtOH.
The amine mixture obtained from the degradation of quaternary salts with 2 or 3 ethyl groups has to be distilled through a 50 cm rotationstring column (Note: I think that such a device is definitely out of reach for the aspiring hobbyist.) The lower boiling amines are distilled off and weighed. The weight of the second fraction can be determined via calculating the difference.
Example: 22.9 g (0.1 mol) dimethyl-diethyl-ammonium iodide are heated for 4 hours in 61 g (1 mol) ethanolamine.
Obtained are 7.310 g of amine mixture AB. Separation:
A: 0.510 g (0.480 g) = 9.5% dimethyl-ethyl-amine; B: 6.90 g (6.65 g) = 90% methyl-diethyl-amine.
A: boiling point 37.5 °C (lit.: 37.5-39 °C); B: boiling point 65¬–66 °C (lit.: 66 °C).
Speed of reaction: After 4 h were 82% of the employed salt degraded.

Degradation of quaternary anilinium salts
Speed of reaction: 3 weighed samples of 0.005 mol trimethyl- or triethyl-phenyl-ammonium iodide each are mixed in 0.25 mol ethanolamine per sample in a reagent tube and heated for 11, 16 and 20 min, respectively, in a bath at 100 or 150 °C. With stirring the salt dissolves, meanwhile a second, liquid phase starts to separate. The reaction is stopped by the addition of 5 ml water, extracted with 5 ml petrol ether and the remaining residues finally weighed out after the solvent has been removed with the aid of vacuum. For exemplary results see Table 6. The products of the cleavage can be checked by converting them back into the starting salts (…again with methyl iodide in MeOH).

Preparative cleavages
Trimethyl-phenyl-ammonium iodide: 13.5 (0.05 mol) salt are heated under reflux for 5 min in 16 g (0.25 mol) ethanolamine. The salt already dissolves quantitatively before the boiling point is reached. After cooling down, 50 ml water are added and the atop floating layer of dimethylaniline is extracted with 3x30 ml petrol ether. After drying the extract over sodium sulphate and removal of the solvent are obtained: 6.0 g (99%) dimethylaniline, boiling point 192–193 °C. Iodmethylate: melting point 217–218 °C (lit.: 217.5–218.5 °C).


The second Chemische Berichte-paper deals with the application of the aforementioned procedure to cyclic quaternary amines, like pyridiniums and alike. The relevant details in short:

    * Dimethyl-pyrrolidinium iodide gets cleaved quantitatively after 6 h boiling in ethanolamine, producing 50% N-methyl pyrrolidine and 20 % dimethylamine. The latter originates from the heterocycle, ie. a large fraction of the pyrrolidine gets destroyed.
    * Methyl-ethyl-pyrrolidinium iodide, same treatment. Due to the lower reaction speed of the ethyl residue will the fraction of ring-destruction increase further to ca. 58%.
    * Diethyl-pyrrolidinium iodide, same conditions as before. By introducing a second ethyl the ring-cleavage becomes the predominant reaction. Destruction of the ring occurs quantitatively.

    * Dimethyl-piperidinium iodide in boiling ethanolamine doesn't give a fuck and reacts rather without any hassle to the expected N-methyl-piperidine; 96% yield.
    * Methyl-ethyl-piperidinium iodide gives only mono-substituted piperidine, too, with the ethyl-isomer being produced in 90% yield (the remaining part is the N-methyl isomer).
    * Diethyl-piperidinium iodide needs just a bit longer (twice the time) due to the lower reactivity of the ethyl chains, but ring destruction is practically not observed. Tough lil' critter!

    * Quaternary salts of hexamethylene imine behave like the piperidinium congeners.

    * Tetrahydro-isoquinolinium salts behave practically like the benzene-annelated piperidines; 2 ethyls take longer to react than 2 methyls, but no ring cleavage takes place; everything else smooth and nice as before.

    * Trying the same with a morphinane-salt, VII, surprisingly led only to product VIII! The phenethylamine-moiety within the morphinane-skeleton behaves like the beta-vinyl in the previous paper, where fast and selective cleavage of a butenyl in presence of ethyl-residues was observed. Hence, the desired N-alkyl-switch fails with this kind of heterocycle. What a pity...

    * The experiments with quaternary tropine-salts were, although not really a surprise, disappointing, too. Tropine combines a pyrrolidine and piperidine ring system, and ring cleavage was therefore not entirely unexpected. Tropine iodmethylate and pseudotropine bromomethylate yielded only 25-35% of the expected base, whose structure was verified by reaction back into the quaternary salt and subsequent IR-spectroscopy. But already from N-propyl-nortropine bromomethylate or pseudotropine-bromo-N-butylate one obtains only wild mixtures of hardly identifiable fragments, all obviously stemming from ring cleavage.


While this procedure is suitable for switching the N-alkyl-residues on pyridines and anilines, it appears impractical for morphinanes and tropine-salts.

The experimental procedures for the ring-derived salts are essentially the same as in the previous paper. Only difference seems to be that the desired products are not always volatile and may remain in the reaction mixture (...instead of distilling over). See for example the run with dimethyl-pyrrolidine:
30.5 g distilled ethanolamine (0.5 mol) and 22.7 g dimethyl-pyrrolidinium iodide (0.1 mol) are heated under reflux for 6.5 h in an apparatus like described above (round-bottom flask, high setup to allow sufficient reflux, Liebig condenser and well chilled receiving flask on the other end). Volatiles are trapped in the receiver (A), non-volatiles remain in the reaction solution (B).
Attaching a small distillation bridge to the flask in which A was collected allows separation of highly volatile (A1) from the less volatile amines (A2). A1 proved to be dimethylamine (FUMEHOOD!) in ca. 20% yield. A2 was distilled with a little bit more force (ca. 80 °C) from solid KOH and proved to be N-methyl pyrrolidine; 51% yield.
To the remaining reaction mix B are added 5.6 g KOH (0.1 mol) in 100 mL absolute EtOH and the alcohol then distilled off at 100 Torr. The residue (B1) is extracted with 5x60 mL ether, the combined ether fractions dried over sodium sulphate and by fractionating distillation separated. The small forerun (12 Torr, 65-75 °C) contains 3.4 g ethanolamine, the main fraction (12 Torr, 137-139 °C) 3.9 g N,N-dimethyl-N'-(?-hydroxyethyl)tetramethylene diamine, IV, in 24% yield. [...we continue here in a second...]

The scheme on page 404 in the second Chemische Berichte-paper shows the presumable fragmentation of the pyrrolidine-ring. The side-product V, N-(?-hydroxyethyl)pyrrolidine, can not be obtained by distillation, because the boiling points of V, ethanolamine and methylethanolamine (the expected side-product of the dequaternization procedure!) are too similar. The authors describe an alternative approach: After neutralisation with HI, diamine IV can be obtained by ether-extraction. By adding excess acrylonitrile (!!!) it is possible to quench primary and secondary amines, while alcohols don't react under these conditions. Thus, the irrelevant bases are derivatized into high-boiling compounds, from which V can be obtained by destillation.

...

Seriously ... acrylonitrile ... is not an option, IMHO. Flammable, toxic, carcinogenic and with an awful smell is acrylonitrile definitely the last option when the synthesis is of highest importance and you have a fumehood available.

[...continuing the work-up from above...] The remaining residue of the destillation (B1) is distilled from potassium iodide (b.p. 55-65 °C, 5 Torr; we call this B2), then combined with the forerun of B1 and with concomitant external cooling treated with 37.0 g freshly distilled acrylonitrile (0.7 mol). Let rest overnight. The next morning distill off first the remaining excess acrylonitrile and the amines, which didn't take part in the reaction. Then, at b.p. 75-81 °C and 12 Torr, isolate 2.9 g of a material, which is again treated with some acrylonitrile (~2.0 g). You already know what's next...distill off...and finally get 2.0 g of compound V.

...what a cluster-fuck!

With the quaternary piperidinium salts the whole process is much easier: 24.1 g (0.1 mol) dimethyl-piperidinium iodide are boiled under reflux for 7 h in 30.5 (0.5 mol) ethanolamine. Collect the distillate in the receiver (over solid KOH) and as soon as the reaction has finished distill just once more directly from the KOH. Yield: 9.5 g (96%) N-methyl piperidine, b.p. 106-107 °C.
« Last Edit: July 11, 2012, 03:21:01 PM by Enkidu »

atara

  • Dominant Queen
  • ****
  • Posts: 256
Re: Demethylation of quaternary ammonium salts
« Reply #2 on: July 11, 2012, 07:22:06 PM »
I keep trying to think of a salt for an ion exchange, but nothing comes to mind. Lead (II) acetate should work, but, y'know, it's lead. Bismuth acetate works too... but who wants to spend hours fucking about with Pepto-Bismol to make a little DMT?

(Lead iodide and bismuth iodide are insoluble in water. So are silver and mercury iodide, but the respective acetates are very reactive)

EDIT: Tin (II) iodide is also not very soluble, so tin (II) acetate should work. Solutions of tin (II) are of course unstable.
« Last Edit: July 11, 2012, 07:27:56 PM by atara »

Enkidu

  • Global Moderator
  • Foundress Queen
  • *****
  • Posts: 677
Re: Demethylation of quaternary ammonium salts
« Reply #3 on: July 11, 2012, 07:56:14 PM »
You could test the liquid phase for iodide with silver nitrate, since silver acetate is soluble. But I'm not sure how accurate the test would be, the accuracy depending on the concentration of iodide in the liquid phase.

akcom

  • Dominant Queen
  • ****
  • Posts: 430
Re: Demethylation of quaternary ammonium salts
« Reply #4 on: July 12, 2012, 04:59:05 AM »
Why wouldn't sodium acetate work?  I'd imagine that heating with the acetate would lead to the (irreversible) formation of the methyl acetate and sodium iodide.

carl_nnabis

  • Subordinate Wasp
  • ***
  • Posts: 245
Re: Demethylation of quaternary ammonium salts
« Reply #5 on: July 12, 2012, 05:35:24 AM »
to bee pendantic (sorry ;D), actually methyl acetate formation is reversible (base catalyzed ester hydrolysis)
"It's like the drug trip I saw when I was on that drug trip!"

atara

  • Dominant Queen
  • ****
  • Posts: 256
Re: Demethylation of quaternary ammonium salts
« Reply #6 on: July 13, 2012, 01:02:39 AM »
Why wouldn't sodium acetate work?  I'd imagine that heating with the acetate would lead to the (irreversible) formation of the methyl acetate and sodium iodide.

It could, actually. I don't think there's any reason to expect iodide would interfere with the reaction.

akcom

  • Dominant Queen
  • ****
  • Posts: 430
Re: Demethylation of quaternary ammonium salts
« Reply #7 on: January 12, 2013, 06:33:15 AM »
to bee pendantic (sorry ;D), actually methyl acetate formation is reversible (base catalyzed ester hydrolysis)
A tertiary amine will not be basic enough to catalyze the hydrolysis of methyl acetate.  In this case, the formation of methyl acetate is non-reversible.