Author Topic: A better way to aminate bromosafrole?  (Read 3222 times)

0 Members and 1 Guest are viewing this topic.

ning

  • Guest
A better way to aminate bromosafrole?
« on: April 02, 2004, 11:34:00 AM »
As we know, the bromosafrole route to MDMA is very attractive in its simplicity. Unfortunately, it also has certain flaws that render it less attractive than it might seem at first glance. Foremost among these flaws is the crappiness of the amination step. You need to use a huge excess of methylamine (which is not something you want to throw around like that), and typically a bomb is called for, because the boiling point of methylamine and solubility at reasonable temperatures is, well, low.

One way to overcome this problem is to use a different reagent, one that doesn't overalkylate. Azide, for example. But that's annoying and hard to get, and anyway, it'll only get you MDA from safrole. We need a better way.
Here is my highly theoretical suggestion, as a method I haven't seen any mention of at the hive:













Molecule:

benzalimine route ("c1cccccc1C=NC.c2ccccc2CC(C)Br>>c1ccccc1C=[N+](C)C(C)Cc2ccccc2.[Br-]")



This little nugget seems very interesting. I know there's at least one reference to it on Rhodium's page(

https://www.thevespiary.org/rhodium/Rhodium/chemistry/amphetamine.methylation.html

), but only for methylation...

How it works (apologies to those who already know):

Basically, a primary amine (i.e. methylamine) is stirred with benzaldehyde and dehydrated to form an imine. While other aldehydes would work, benzaldehyde is particularly preferred because it forms a stable adduct with the amine because the double bond of the resulting imine conjugates with the benzene ring, and it doesn't undergo self-aldol condensations. Once this imine adduct is formed, a stoichiometric amount of some alkylating agent is added, for example ethyl bromide, methyl iodide, or dimethyl sulfate, and the mixture is heated. Since the amine has three slots already filled up, only one more alkylation can occur, giving a quaternary iminium salt. This salt is then hydrolyzed to regenerate the benzaldehyde and give pure secondary amine. (i.e. MDMA) You could say the benzaldehyde acts as a blocker, to prevent overalkylation.

This is a very exciting general possibility, because it completely eliminates the overalkylation that would otherwise occur. It also has three other not-so-obvious advantages.
- First, the benzalimine has a VERY high boiling point, so you can heat that fucker to kingdom come without a bomb. This should speed the reaction up considerably.
- Secondly, unlike a normal alkylation reaction where HBr is evolved by displacement of the bromide by the amine, this reaction generates NOTHING. The two compounds come together and stay that way until hydrolyzed. Very handy, because there's no equilibriums to displace, no HX acids/gasses to sequester with pyridine or further excesses of amine.
- Finally, both compounds (the benzalimine and bromoalkane) are oil soluble, while the quaternary compound formed is much less so. I suspect that this will tend to help the reaction run to completion by removing the product from the mixture. If you ran it in boiling xylene, or mineral oil perhaps, I guess you could see crystals falling out of solution.

MMMMM, sounds yummy! Stoichiometric or 2x excess of amine, high yields, fast reaction, and the benzaldehyde is recyclable! That's what I like to hear! Only one problem--nobody seems to have done this reaction with a secondary haloalkane! In fact, most lit refs (and there aren't many) seem to copy the Org Syn procedures verbatim! Geez, these chemists sure are a stodgy bunch, aren't they?!
Never bothering to investigate something cool and OTC just because they already have 5 perfectly good ways of doing things already! Well, that's what we have the hive, I suppose, so we'll just have to see about such things ourselves!

So, suitably pumped up about this seeming diamond in the rough, I did some lit searching and paper abstracting, and here's what I found.

First, the procedure mentioned on Rhodium's page--N-methylation of amphetamine with dimethyl sulfate.
"A Rapid, Convenient Preparative Procedure for Phenethylamines", JMC 1972, 214

p-Methoxyphenethylamine, generated from 100 g (0.536 mole) of the hydrochloride by stirring with concentrated aq NaOH, was treated with 100 ml benzene and 70 g (0.66 mole) of benzaldehyde. A mildly exothermic reaction began at once. The mixture was heated under reflux until no more water was present in the condensate (circa 1 hour), then, without cooling, an attached Dean-Stark trap was removed and a solution of 82 g (0.65 mole) of DMS in 200 ml of benzene was added through the condenser at such a rate as to maintain reflux (15 min). The 2-phase mixture was heated for 90 min on the steam bath, cooled slightly, treated with 200 ml water, and heated for an additional 20 min. After cooling in ice, the aqueous layer was washed twice with ether to remove unreacted benzaldehyde and made strongly basic with 50% NaOH. Two ether extracts of the basic aqueous phase were added to the amine layer which separated, and the resulting solution was evaporated under aspirator vacuum for 30 min, leaving 90 g (102%) crude N-methyl-p-methoxyphenethylamine. This material was dissolved in 500 ml of 20% anhydrous ethanol-ether and treated with 50 ml of concentrated HCl with swirling and cooling to yield the white, crystalline hydrochloride, which was washed thoroughly with ice-cold ethanol-ether and dried. Yield: 83 g (77%)

JOC 1989, 2448
: Methyl(trideuteriomethyl)amine

N-benzylidenemethylamine (3.58 g, 30 mmol) and CD3I (5 g, 34 mmol) were pipetted into a small (40 ml) stainless steel pressure cylinder, sealed, and heated to 100 C for 24 h. After cooling, to 45 C (not lower, to prevent the contents from solidifying), the cylinder was opened cautiously in a fume hood and the contents were poured into a 100-ml flask, along with the rinsings of the cylinder (10% HCl, twice 5 ml). The reddish brown slurry was refluxed for 2 h and the resultant mixture placed in an addition funnel and dropped slowly onto NaOH pellets to liberate gaseous Me(CD3)NH. [...] An 83% yield of Me(CD3)NH.HCl was obtained.

Here's a fun one--they use methyl BROMIDE for the alkylation...freaky.

JACS 1992, 5091
: n-hexadecylmethylamine

To a solution of n-hexadecylamine (3.0 g, 12 mmol) in 100 ml toluene was added 1.32 g of benzaldehyde (12 mmol). This solution was heated at reflux for 2 h and the water formed during the reaction removed by means of a Dean-Stark trap. After the solution cooled at room temperature, the organic solvent was distilled under reduced pressure, leaving quantitatively the imine derivative. This material (4.02 g, 12 mmol) dissolved in 100 ml of methylene chloride was poured in a screw-top pressure tube and cooled to -78 C. After the addition of 3.3 ml of methyl bromide (60 mmol), the tube was sealed and kept at 70 C in an oil bath for 48 h. The solvent was then evaporated and the residue treated with 100 ml of 0.1 M NaOH and extracted with chloroform. The evaporation of the dried organic layer (Na2SO4) afforded a crude material that was purified by dissolving it in acetone and precipitating the HCl salt of the amine upon addition of concentrated HCl. The white solid n-hexadecylmethylamine HCl, 3.18 g (91%), was collected and washed several times with acetone.

Also, please see Org Syn CV 5, 736 and 758, N-methylbutylamine and N-methylethylamine.

Finally, there is a paper in French which does the same thing. Maybe someone skilled in the language could translate the relevant parts--Journal of Fluorine Chemistry 93 (1999), 145
They seem to also reduce imines to amines with sodium dithionite in there. Perhaps this too could be useful? Hmmm...


So what's the point? The point, my friends, is this: If you can alkylate amphetamine with methyl iodide, or some bromide in high yield by this method, why could you not instead alkylate methylamine with bromosafrole or 2-bromo phenylpropane?

I have seen:
: all kinds of things alkylated with methyl iodide
: something alkylated with alkyl bromides (somewhere...)
All in high yields!

Yes, I'm sure it would be slower, but the steric hindrance is roughly the same, and you wouldn't need sealed tube or anything because the boiling point of both components would be high enough. Just a little dry acetone or DMSO, a pinch of NaI perhaps, and away you go! I wouldn't be too surprised if yields in excess of 70% could be realized, without needing the vast excesses of methylamine that haunt the normal bromosafrole method. Probably a 2x would be more than enough.

Thoughts?


Rhodium

  • Guest
Looks better on paper than in the flask...
« Reply #1 on: April 02, 2004, 12:14:00 PM »
Yes, I'm sure it would be slower, but the steric hindrance is roughly the same

There is a significant difference in sterical hindrance, as well as intrinsic reactivity between a secondary alkyl bromide and a primary alkyl iodide! Not to mention that methylamine benzaldimine being less stable than amphetamine benzaldimine (less possible resonance forms).

I'm not saying it would be impossible to get any yield at all, but definitely that this suggestion offers no advantage whatsoever over the two-step route via safrole azide followed by reduction using any suitable route.


ning

  • Guest
There is a significant difference in sterical...
« Reply #2 on: April 02, 2004, 03:40:00 PM »
There is a significant difference in sterical hindrance, as well as intrinsic reactivity between a secondary alkyl bromide and a primary alkyl iodide!

Absolutely. But how much difference, I wonder. They used a primary bromide, and it worked.

methylamine benzaldimine being less stable than amphetamine benzaldimine

Doesn't seem to make any difference. See the second procedure in my post, where they use just that, and get fine yields.

this suggestion offers no advantage whatsoever

None? It sounds alot more OTC. No reduction required. No azide salts required. Maybe you have discretionary access to these things, but then, if everybody did, we wouldn't need the hive, would we?

Actually, I don't see why you so readily dismiss this idea. It's not so different from the Delepine reaction!

I would appreciate it if, before squashing something new, you could give it a try. Perhaps this is too much to ask?
I will continue to trawl beilstein looking for more informational references than these that I found.


Rhodium

  • Guest
The Delepine isn't too hot either...
« Reply #3 on: April 02, 2004, 03:48:00 PM »
Doesn't seem to make any difference. See the second procedure in my post, where they use just that, and get fine yields.

Yes, with methyl iodide. It is not any specific part of the proposal which causes the idea to break apart or anything, but there is a lot of disadvantageous details which together adds up to something I believe will cause problems.

Actually, I don't see why you so readily dismiss this idea. It's not so different from the Delepine reaction!

That is not to the advantage of this proposition. I have yet to see any reference to any really successful Delepine reaction being performed on a secondary halide. See the yields and reaction times in

https://www.thevespiary.org/rhodium/Rhodium/chemistry/delepine.txt



I thought I expressed myself rather gentle - I pointed out possible problems, but I did also say I didn't believe it was impossible to do - just that the reaction wouldn't be more high-yielding or easier than the azide route. You did find a possible advantage I didn't think of - the OTC-ness in certain parts of the world (where I live, sodium azide is available from pyrotechnics suppliers...).


Vitus_Verdegast

  • Guest
otc-ness of azide
« Reply #4 on: April 03, 2004, 12:58:00 AM »
I think sodium azide is available OTC practically everywhere, in airbags. They contain quite a lot, and the compound is easily isolated from the material inside (contains also KNO3 and SiO2 if I recall correctly). UTFSE for it.


ning

  • Guest
Sigh....
« Reply #5 on: April 04, 2004, 02:05:00 PM »
True, Rhodium, true. I'm going to do some more study to prove to myself whether this method is feasible or not.

Vitus, I know about the azide, though I'm not so sure how OTC it is for me. (Maybe it should be called ITJ rather than OTC--In The Junkyard) My point is actually that AFAIK, the azide route can only give you primary amines (i.e. MDA). That is also the case with the Delepine reaction. If you want secondary amines, you will have to do something else. If this process (I think it's called the Decker method) works well enough, it would allow the arbitrary mating of of any primary amine with any alkyl halide in reasonable yield. For me that makes it something worth thinking about, even though there are some significant issues to be overcome.

Actually, I'm somewhat enamored of Rhodiums idea for reducing isocyanates of late. I've heard of something like that before in the literature. Hmmm.


ning

  • Guest
Here's what I found
« Reply #6 on: April 04, 2004, 10:45:00 PM »
A pile of hints, and perhaps an order-of-magnitude idea of the reaction time to be expected.

I searched the web, Beilstein, and Organic Syntheses for reactions involving the alkylation of benzyl alkyl amines with secondary alkyl halides. My beilstein query looked like this:

Bn.N.R1 + R2.CHX.R3 --> Bn(R1)N.CH.(R2)R3

Where R1 = ALK and ALH, R2 = R3 = G

I found about 12 decent refs, and got 4 of them.

JOC 1962, 1542 has a few reactions. They make quaternary ammonium salts by reacting benzyl methyl amines with various alkyl halides in acetonitrile. The reactions of interest are the following:

- Benzyl dimethyl amine + isopropyl iodide --> benzyl dimethyl isopropyl ammonium iodide
standing, one day. (94%)
- Benzyl dimethyl amine + isobutyl bromide --> benzyl dimethyl isobutyl ammonium bromide
reflux, two days. (88%)

JOC 1995, 6776 reacts a big, bulky alpha-carbonyl bromide with benzyl methyl amine in HMPA. They also did DMF and DMSO with similarly good yields (>94%), but they were doing some stereoselective bullshit and HMPA gave the best e.e. This has to be seen to be understood, so here it is:












Molecule:

ugh ("c1ccccc1CNC.C(C)(C)(C)OC(=O)C1CN(C)C(=O)N1C(=O)C(C)Br>>C(C)(C)(C)OC(=O)C1CN(C)C(=O)N1C(=O)C(C)N(C)Cc1ccccc1")


that's a real monster, ain't it? 14 hours stirred in HMPA at 25 C yielded 98%.

JMC 1966, 88 does the same sort of thing, only it's a little smaller. Interestingly, they also perform the mannich reaction here (getting medianly crappy yields ~54%). But that's not what we're interested in right now. All of their compounds are of the form R-SO2-NH-Ph-COCH(R)-Br, reacted with benzyl methyl amine. We'll just ignore the first R, since it doesn't really matter. They react in acetonitrile.

- R = Me, 16 hours, yield 24%
- R = Me, 16 hours, yield 29%
- R = Me, 8 hours, yield 48%
- R = Et, 30 min, yield 57%

Yields consistently sucked, but interestingly, yields seemed to improve as the first R group got bigger, providing more and more bulk. This suggests that overalkylation may have been the yield-killer.

In Tet 1986, 2117 they're making phenmetrazine-ish dopamine agonists! Nifty! It's a small world after all. 8)  The reaction of interest is again an alpha-carbonyl bromide attacking benzyl methyl amine.
The reaction looks like this:
PhCOCH(Me).Br + BnNMe --> PhCOCH(Me)N(Me)Bn

This is cool, because it's a phenethylamine (A PMMA analog, actually), just with a carbonyl group alpha to the bromine. I guess this activates it and makes it a better leaving group, but all the steric factors should be the same here as in the real thing, so it's cool to see this. Since this is just an alkylation to tertiary and not to quaternary, they use catalytic KI in acetone. The solution is stirred at room temperature for 3 days.

- PhCOCH(Me).Br + Bn2.NH : 70%
- PhCOCH(Me).Br + Bn(Me).NH : The bastards don't say!

But we can imagine that a smaller amine is going to alkylate better than the bigger one.

Other papers which may be worth investigating, if someone could get them:

J. Pharm. Sci 1981, 699 : Bn(Me)NH + RCOCH(Me).Cl
Liebigs Ann. Org. Bioorg. Chem. 1995, 211 : Bn(Me)NH + iPr.Br
Chem. Ber. 1912, 1306 : Bn(Et)NH + iPr.I
Chem. Pharm. Bull. 1967, 1294 : Bn(Me)NH + MeCOCH(Et).Cl


The basic assumption behind my searching was that a double bond would act like a single bond for the purposes of the SN2 reaction that must occur for the bromo-compound to attach itself to the amine. I don't actually know if this is valid. Is it? What difference is there between a double and single bond for SN2 purposes, other than slightly shorter bond length and one less hydrogen attached? And how much does a neighboring carbonyl group accelerate SN2 reaction? I would guess it increases leaving group speed quite a bit, but how much? Is there any simple relationship between pKa and leaving rate? Is a chlorine alpha to a carbonyl group roughly as good as an iodine as a leaving group, or what?

Awaiting the comments of those with real chemistry skills ;)


ning

  • Guest
Anybee know what's the difference...
« Reply #7 on: April 08, 2004, 10:44:00 AM »
in rate and steric hinderance, between this












Molecule:

two bonds ("c1ccccc1C=NC.BrC(C)(C)>>c1ccccc1C=[N+](C)C(C)C.[Br-]")


and this?












Molecule:

one bond ("c1ccccc1CNC.BrC(C)(C)>>c1ccccc1CN(C)C(C)C")



Any leads, thought? Ning wants to know.


Lilienthal

  • Guest
hint: planarity, free rotation, fixed bonds...
« Reply #8 on: April 08, 2004, 11:00:00 AM »
hint: planarity, free rotation, fixed bonds...

ning

  • Guest
Hint, hint...
« Reply #9 on: April 11, 2004, 05:47:00 PM »
I suppose that means "double bond is more hindered". Which wouldn't bee too surprising.
But what I would really like is some reference where that effect is quantified. Got any?