Delepine rxn - an important insight?

[Antoncho]

Dear Bees!


Most of you probably have heard of Delepine rxn - the possibility to react some haloalkane (say, bromo/iodosafrole) with hexamethylenetetramine (aka HMTA aka camping fuel tablets) to get primary amines. This trick has been regarded as especially seductive since direct amination of the interesting haloalkanes proceeds satisfactorily only with methylamine, while for NH3 (which is a weaker base) even such conditions as heating in a bomb give unsatisfactory (~30%, IIRC) results. To circumvent this obstacle people have resorted to roundabout routes - sucn as haloid/azide exchange, followed by reduction.

The main reason of Delepine rxn being used very little seems to bee the inconsistency of the experimental results. As one of the maitres (Os, was it you?) put it "...some people succeed with it, some not. Mostly not."

Tonight I was reading an AMAZING Russian chemistry book called - well, you don't care anyway . Here's what it says on Delepine:


ArCH₂X ^__> [ArCH₂(CH₂)₆⁺N₄]X^{- __}> ArCH₂NH₃⁺X^-

This synth represents the 1st part of Sommelet rxn. If one carries out the hydrolysis of the formed haloid/HMTA complex with the mixture of EtOH and conc. HCl, it is possible to arrive at the primary amine as the main product.

Here's an example:

Beta-alanine.

2,74g NaHCO3 is added to a soln of 5g b-bromopropionic acid in 15mls water and 10mls EtOH.To thus obtained neutral solution there's added 4,57g HMTA and the mixtr is allowed to stand for 15hrs. The 50mls EtOH is added until slight turbidity; scratching the beaker's side w/a glass rod leads to massive precipitation of the addition complex. On cooling/filtering  twice one obtains 9,5g grams of the xtals.

Thus obtained complex is refluxed for 15 hrs in 120mls EtOH/15mls conc.HCl. The yield - 85%.

Another example includes hydrolysis of 1-iodoheptyne by stirring for 3 days in ethanolic HCl.

Yet another example (2-bromoallylamine) describes preparation of the addition complex by refluxing the reagents in chloroform with subsequent hydrolysis of the precipitated complex . This one particularly drew my attention since it is the method reported to work well (although, not on an amphetamine but on a methcathinone) by a Russian chemist of the past named Assholium - unfortunately, he didn't give any details.



As you can see, the amine salt isn't formed at once - more so, the hydrolysis of the HMTA/haloid complex requires some significant effort and time. Now let's analyze the procedure know to The Hive thus far (a quote from

https://www.thevespiary.org/rhodium/Rhodium/chemistry/dmmda-2.html

- which, i assume, is the document used as a guideline for the most experimenters on this subject. I'll call it Doc#1 later on)) :

Put the undistilled iodo-dillapiole in 10x amount of alcohol in a 500ml RB flask, heat and begin adding slight molar excess of hexamine until all is in. Refluxing this for 6 hours gave an eventual yield of DMMDA-2 which is quite rewarding. No one should doubt the power of delepine!

The workup:

Add HCl until the ph is 2, and filter off the ass-load of formed ammonium chloride.

The red liquid is evaporated until a mass of hard rocklike crystals remained. This is impure DMMDA-2. The crystals CANNOT be cleaned, with ether or any other washes. Freebase and distill to give clear, off colored DMMDA-2.
The yield is 50% from the beginning. SWIM got ~12 g at the end.

Compare this to the original article coming from

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

(Doc#2 further on):

To a solution of 1 mole of hexamethylenetetramine in eight to ten times its weight of hot 95% alcohol slightly more than 1 mole of sodium iodide is added. One mole of the alkyl chloride or bromide is then added and the solution is allowed to stand until the precipitation is complete. The period of precipitation will vary from a few minutes to several weeks depending upon the substance. The longer the carbon chain the longer is the time required for the precipitation. The mixture containing the precipitate is saturated with hydrogen chloride gas, whereupon the precipitate dissolves and ammonium chloride precipitates. When the latter has been filtered off and the alcohol removed by distillation, the resultant impure hydrochloride is converted into the pure amine by distillation with an excess of sodium hydroxide. (Note: this doc also contains details on performing the rxn in chloroform).

As you can see, there isn't a single mention of any 'hydrolysis stage' in Doc#1 (I'd guess that - MAYBEE - the author may simply have left the acidified rxn mixture overnight w/out mentioning it in his report).
And the 'shitload of ammonium chloride' described by Egotrip is probably the unhydrolysed addition complex (it is rather improbable that NH4Cl would've been crushed from ethanolic solution by aq. HCl - however, OTOH, the HMTA/haloid adducts seem to bee more soluble in water than in non-polars).

While ethanolic HCl seems to do the job rather fast (at least, on primary alkylamines), another example from Doc#2 calls for 2h reflux in 300ml EtOH/20mls conc. HCl.

Also in Doc#1 there's no mention of any precipitate in the 1st stage (or was there too little of it to xtallize out?). The truth is that the addition complex isn't very soluble in near anything: neither in EtOH, nor in CHCl3, nor in water (as can bee seen from earlier in Doc#2). And thus, MAYBEE, it may serve as an excellent indicator of the rxn progress.




Those were the facts. The conclusions i draw from all of this are the following:

1) It is desirable to carry out the reaction of haloalkane with HMTA in some solvent that would allow precipitation of the adduct (this will take experimentation since we don't know the exact characteristics of our adduct in this case, however we have good guidelines). Concentrating the solution followed by some cooling/scratching might well bee the perfect way to go.

2) Complete hydrolysis of the adduct should bee ensured bee4 final A/B xtraction. Maybee, gassing of the EtOH is worth trying - who knows? In any case, i'd guess that some refluxing won't hurt.


So, bees. Maybee i'm just saying things that were obvious to you, but nevertheless i found this to bee worth your attention - please, let me know what you think of this.


Antoncho

[PolytheneSam]

I read somewhere that when making benzylamine from benzyl chloride the Delepine or Gabriel reaction is used because if you try to react benzyl chloride with NH3 you get a lot of secondary and tertiary amines because benzyl chloride is so reactive.

http://www.geocities.com/dritte123/PSPF.html

The hardest thing to explain is the obvious

[mercvap01]

When you said that you used NH3/iodosafrole in a bomb and only got 30% yield did you use ammonium hydroxide or anhydrous ammonia?  I realize that the water in the ammonium hydroxide messes the reaction but thought the anhydrous should work.  Also what about using ammonium acetate instead of hexamine in the delepine?

[Rhodium]

mercvap: The bomb method uses a solution of ammonia in alcohol, neither aqueous ammonia or free anhydrous ammonia is a good idea. And no, in the Delepine reaction you can't substitute the hexamine with anything else.

[SPISSHAK]

{To a solution of 1 mole of hexamethylenetetramine in eight to ten times its weight of hot 95% alcohol slightly more than 1 mole of sodium iodide is added. One mole of the alkyl chloride or bromide is then added and the solution is allowed to stand until the precipitation is complete. The period of precipitation will vary from a few minutes to several weeks depending upon the substance. The longer the carbon chain the longer is the time required for the precipitation. The mixture containing the precipitate is saturated with hydrogen chloride gas, whereupon the precipitate dissolves and ammonium chloride precipitates. When the latter has been filtered off and the alcohol removed by distillation, the resultant impure hydrochloride is converted into the pure amine by distillation with an excess of sodium hydroxide.}

This procedure looks to be a mess I see he's trying to perform a finkelstein metathesis reaction with Sodium Iodide and the alkyl X compound at the same time he's forming the hexamethylenetetramine complex (delphine reaction), but definte good point on the fact that hydrolysis of the intermediate is not mentioned in either example. Not to mention that the proposed finkelstein reaction will not work on anything but a primary alkyl halide, that's something also overlooked, i pondered if it could be done say a finkelstein reaction involving chlorosafrole, Sodium Iodide in acetone and chloro safrole is a secondary alkyl halide so according to Vogel (I think) that will not work.

[Rhodium]

Adding a catalytic amount of NaI/KI to a solution when doing any kind of nucleophilic substitution with less reactive alkyl halides is common laboratory practice, so that is nothing strange.

Also, the Finkelstein swap will definitely work on secondary chlorides/bromides, so if you have made chlorosafrole (PTC/37% HCl), refluxing it in NaI/acetone will without doubt produce iodosafrole.

[Osmium]

The delepine might be all fine and dandy for primary halides, but I haven't seen many examples where it was successfully used for secondary halides.
The organic 'tail' of our halide is rather big, which according to Antonchos post means reaction will take quite some time even with primary halides.
Now add the slowing down effect of using a secondary halide and I really don't see much success for the experimenter.
I'm not fat just horizontally disproportionate.

[SPISSHAK]

HMTA is a rather bulky molecule, the steric factors may preclude this from working no?

[Rhodium]

Yes, it is a very strong argument - several articles I have seen state that secondary halides are unreactive towards HMTA, and only scattered reports of secondaries actually yielding amines (a few in the literature and a few from bromosafrole cooks). I have not seen any conclusive evidence either way though, mostly just footnotes and hearsay.

[SPISSHAK]

The only thing you could do in that case is extend the reaction times greatly, bump up temperatures maybe, only expirimentation will prove this.

[mercvap01]

In doing the Delepine rxn if you are starting with iodosafrole you would not need to use sodium iodide because I thought the only reason it is used with chloro and bromo compounds was because it converted the molecule to a iodide in situ.  From my understanding iodosafrole should be the most reactive and even though it is secondary mabe it will react slowly like you said due to steric hinderance.  I would be afraid to use too much heat though especially over a long period of time I would guess it would break up the methylenedioxy ring.  Anyway I would be interested to know if anybody ever got this to work via iodosafrole.  A friend of mine had good success with converting iodosafrole to MDMA using methylamine in a bomb but only got 50% yield and I am more interested in MDA>

[SPISSHAK]

Iodosafrole?  just curious. and I thought iodosafrole was prone to elimination HX at elevated temperatures? How did he get any yeild under "pipe bomb" conditions?

[ChemisTris]

Haloarylation of unsaturated compounds by aromatic diazo compounds. X. Synthesis of DL-phenylalanine and its homologs.
Yurkevich, A. M.; Dombrovskii, A. V.; Terent'ev, A. P. State Univ.,  Moscow, Zhur. Obshchei Khim.  (1958),  28  227-30.
CA 52:72206

Abstract:
cf. C.A. 51, 15454g; 52, 9019i. 
Heating 1.5 g. PhCH2CHClCO2H in an ampul with 5 ml. 25% NH4OH and 2 g. (NH4)2CO3 60 hrs. at 45°, followed by refluxing 10 hrs., gave 40% phenylalanine, decomp. 271-2°; a similar reaction using liquid NH3 with NH4Cl added for catalytic effect gave a 42% yield in 3 days at room temp. PhCH2CHBrCO2H and NH4OH gave 74% yield in 7 days at room temp., while an 80% yield was attained with liquid NH3 in 3 days.  p-MeOC6H4CH2CHClCO2H and liquid NH3 in 3 days at room temp. gave 70% p-MeOCH2CH(NH2)CO2H, decomp. 291-3°. 

Heating 4.4 g. p-MeOC6H4CH2CHBrCN with 10 ml. 85% HCO2H and 10 ml. concd. HCl to 100° and treating of the crude product 6 days with NH4OH at room temp. gave 45% p-MeOC6H4CH2CH(NH2)CO2H.  This treated with red P-HI in Ac2O 4 hrs. at reflux gave 90% tyrosine. 

p-ClC6H4CH2CHClCO2H and concd. NH4OH in 7 days at room temp. gave 30% p-chlorophenylalanine, decomp. 239°; the yield being 83% from a reaction in liquid NH3 in presence of NH4Cl.  Similarly was obtained 85% p-bromophenylalanine, decomp. 254-5°.  Similarly formed 96% 2,4-dichlorophenylalanine, decomp. 237-9°.
 
p-O2NC6H4CH2CHClCO2H with NH4OH gave the unsatd. acid under various conditions tried, but if the chloro acid is refluxed in dioxane with urotropine 4 hrs. and the product refluxed with alc. HCl 2 hrs. there is formed 72% p-nitrophenylalanine, decomp. 240-5°.  This method also gave: phenylalanine Et ester HCl salt, 30%, m. 126-7°; p-nitrophenylalanine Et ester HCl salt, 40%, m. 177-80°; PhCH2CMe(NH2)CO2H, HCl salt, 65%, decomp. 236°; 70% p-O2NC6H4CH2CMe(NH2)CO2H HCl salt, m. 273-4°.  Heating 5 g. PhCH2CHClCN and 9.8 g. PhNH2 8 hrs. at 140° gave after an aq. treatment and steam distn. of PhNH2, 41% PhCH2CH(NHPh)CO2H, m. 172-3°.
Got democracy?

http://www.dhushara.com/book/multinet/democ/wed.htm