phosphorus triiodide

loki · guinea pig

i was at WD tonight, reading a thread in which there was some speculation on reaction mechanisms of the red phosphorus and iodine reaction system. one thought lead to another (as they are wont) and eventually i was investigating phosphorus triiodide.

from the wikipedia ( http://en.wikipedia.org/wiki/Phosphorus_triiodide ):

loki · guinea pig

oh, just one little appendix: this method would allow one to substitute white phosphorus for red. (i understand white phosphorus can be found in certain types of welding rods).

oh... stochiometry!

2(P) + 3(I₂ ) --> PI₃ (100%)
PI3 + ephedrine --> iodoephedrine + H₃ PO₃ (90%)
2(iodoephedrine) + 2(H₃ PO₃ ) --> methedrine (90%) + 2(H₃ PO₄ )

yield would be in the range of 81% at my estimate, noticably better than the standard of 70%. i might be wrong about the yield of the phosphorous acid reduction.

pH does not seem to be a big factor here. if the amount of rP/I2 used is calibrated to be exactly enough to convert the ephedrine to iodoephedrine, one could alternatively add K₂ PO₃ solution (if i am not mistaken, this could be straight up phytophthora (sp?) (aka collar rot) fungicide solution (which is a mix of KPO₃ and K₂ PO₃ , most likely calibrated to have a pH of about 6, hence the 'mixture of mono and di potassium salts of phosphorous acid)

or alternatively, sodium hypophosphite solution, which would of course be more efficient (but a bit harder to get a hold of)

the point of doing this would be to save the effort of acquiring the difficult to get reagents phosphorus and iodine to any degree more than absolutely neccessary. One only needs one iodine atom per ephedrine molecule to get to the intermediate.

the phosphorous acid produced in the reaction probably would mostly decompose by the time the reaction is completed anyway.

potassium phosphite is easy to get and cheap and an indispensible treatment for a very common plant disease. one can use white phosphorus in the reaction, which means that other sources can be used (eg welding rods). by focussing the reaction on this two step iodination/deiodination one can minimise the amount of phosphorus and iodine needed, and use the easier to get potassium phosphite to finish the reaction off.

a foot note: i was told by a friend once that one can use HCl and KI instead of iodine in the reaction, and while i was dubious about this, i have had this confirmed by a reliable source that it does, to some degree, work (positive bioassay). The fact that this can be done says that having K ions floating around does not interfere with the reaction. This means that if one has a pile of mostly iodoephedrine, if it was refluxed with potassium phosphite (or indeed it seems very likely sodium hypophosphite solution would be equivalent albeit more efficient), it would be converted to methedrine, because clearly neither the iodination nor the reduction are inhibited by it. the microwave method of iodination using KI (it uses sulphuric acid and kaolin) likewise succeeds in iodinating despite the presence of both sulphates and potassium ions... the sulphate is only problematic for the phos-acid reduction (which it reduces to all sorts of nasty sulphurous crap).

one last comment - the cause of the formation of aziridines and the residual iodoephedrine is two causes behind the toxicity of street speed. aziridines require high temperatures to form, but the iodination reaction proceeds nicely at room temperature, the phos acid reduction step probably requires reflux type temperatures, i'm not exactly sure, possibly it may still run at lower temps but just take longer... but i know that the reaction does not occur very well at much less than 70 degrees, it usually runs around 120-130 normally in fact... but this type of temperature is prone to causing phos acids to decompose to phosphine, and can cause the self-reaction which produces the aziritines (more down lower), it starts at 100 and beyond 130 it starts to form potentially dangerous amounts of diphosphine which autoignites. since hypo is added to HI acid to stop it reverting to I2, this reaction obviously happens at colder temperatures, but this is likely to be a slow reaction. in both reactions, i think a nice 70 degree temperature (exactly what you would get from a steam bath) is sufficient (in other words, those neat hotplates on coffee makers which keep the coffee hot are perfect!

regarding the aziridine formation, this can *only* happen under high reaction kinetics, but it is nothing out of the ordinary when you think about it, think about other reactions where methyl iodide is used to methylate an amine - all you do is mix halide and amine together and they form a quaternary iodide, the aziridine is a quaternary iodide... and of course when it reacts with phosphorous acid in the standard reaction, you end up with straight methyl phenyl aziridine. ugh.

oh yea, my last thought was about extracting after the reaction... since the reaction will have naptha sitting on top of it, and at the beginnig will be in fair excess, by the end of the reaction we have a two phase mixture with hot naptha at the top. rather than fuck around, just throw lye into it and let the ---- go up into the hot naptha and extract it directly from there, water and lye, heat it for 10 mins, separate and then dry and gas or titrate or whatever. I think it would be quite neat to evaporate the naptha/methedrine solution down until the ---- starts to precipitate out (as a liquid) and then put it in the freezer and drive all of the ---- out (i couldn't get any sense out of anyone when asking what ---- freebase's mp is but my suspicion is that it forms a solid in freezer temperatures. then one simply decants the cold supernatant naptha solution and one can then convert it to a salt by adding the acid directly and mixing.

loki · guinea pig

it has occurred to me that a solution of lye underneath would be essential for doing the iodination reaction because of the phosphorous acid that is formed would tend to pull ephedrine randomly out of the solution not particularly regarding whether it is iodoephedrine or not, so having sodium ions in a small amount of water underneath would mean that the phosphorous acid formed in the iodination would be neutralised, preventing the undesired movement of freebase ephedrine into the salt as the phosphite.

the ideal would be to put just a little excess above the molar quantity of phosphorous acid expected to be formed, or maybe a little under, and then once the phos acid levels rise past that point it starts pulling the iodoephedrine down into the sodium phosphite solution underneath. i am inclined to prefer to just use lye solution to pull the phosphite out of the reaction and separate it and then use potassium phosphite solution instead and add phosphoric acid to pull down the ephedrine into the aqueous phase.

loki · guinea pig

Heres my proposed method, the quantities mentioned are very small. i know that conventional wisdom says small is hard to get right but the conventional reaction system, to be quite honest, i'm amazed that it works so well when the PI3 is immediately hydrolysed. i am now of the opinion that the short dry cook produces equally good product but is more dependent on purity because the reaction instantly hydrolyses a lot of the PI3 before it gets to iodinate, and extremely dry reactions thus most likely end up with a lot of iodoephedrine due to the heat decomposing the phosphorous acid thus making it impossible for the reaction to complete. (one could theoretically just produce the iodoephedrine and extract it to bioassay to determine the differences of effects.) and one should not forget that in the reaction conditions of such high heat, there is significant chance of the iodide binding to the amine (amines and alkyl iodides bind to form quaternary iodides) and thus producing methyl phenyl aziridine (the phos acid would reduce the iodide). this reaction scheme does both reactions below 90 degrees so should be free of aziridines and because an excess of phosphite is used, there should be no remaining iodoephedrine either.

reaction
1 to a flask add 50ml of naptha, the phosphorus and iodine (about 2g I2 and 1g rP)
2 heat reaction until it appears to stabilise, should produce a distinctly reddish solution
3 add about 10ml of saturated NaOH solution, should be about 3-5g of NaOH in 10ml of water (just pour off the supernatant once it's fully dissolved as much as it is gonna dissolve in 5 minutes)
4 then add the 1.3g of pseudoephedrine freebase and heat the solution for half an hour.
5 add about 10ml of (mono/di) potassium phosphite solution to the mixture and 10ml of ~50% phosphoric acid and heat the solution for an hour.

workup
1 add 50ml of naptha and add NaOH solution until the solution turns milky
2 heat the solution until the naptha is boiling and let it boil 10 minutes then separate, brine wash and put the naptha in the freezer
3 after two hours ---- freebase should have fallen out of the solution (possibly as crystals but maybe just an oil) then decant the cold naptha as much as possible, and if neccessary, allow the remainder of the naptha to evaporate under a fan
4 add conc. HCl to the solution dropwise while on the heat until all the ---- is converted to the salt and then dry it out with a fan and on the heat.

loki · guinea pig

just in case anyone is reading this :/ i am thinking about another synthesis related to this one involving alkylation of the amine with propargyl ( 2-propynyl) bromide.

can anyone tell me if phosphite or hypophosphite could be used to reduce the quaternary iodide that forms when the amine is mixed with the alkyl halide?

i was also thinking of somewhat reducing the legal implications of this synthesis by directly alkylating pseudoephedrine thus avoiding the need for ---- as an intermediate, can anyone see any reason why the addition of the alkyl halide to the amine would not work due to the presence of the benzylic OH?

loki · guinea pig

hmmm i maybe put this thread in the wrong forum? maybe it should be in misc. stims?

anyway, just answering my own question using deduction:

methyl phenyl aziridine is known to form in the rp/i2 reaction as it is commonly done (according to DEA forensic scientists, there's a paper on rhodium about it and the aziridine is mentioned as a reaction byproduct.

therefore, in the conditions, a quaternary iodide as would form when the iodoephedrine is heated hot enough the iodide adds to the amine and binds the carbon to the amine, and the fact that the reaction byproduct is plain methyl phenyl aziridine, this indicates that the conditions of the reaction are such that the halide is reduced from the amine leaving behind the plain aziridine.

in other words, this means that phosphorous acid (which is formed in the rxn from hydrolysis of PI3) is able to reduce a quaternary halide. ergo phosphorous acid would be able to remove the iodine atom. and more than likely a bromine atom too.

diverging once again somewhat, this would tend to indicate that phosphorus acid (or hypophosphorous acid) and possibly even an acidified salt of the acid (ie potassium phosphite plus phosphoric acid) could do this reaction to other amines being alkylated via an alkyl halide. for example, tryptamine --> dmt synth using methiodide as the methylating agent would be efficiently reduced with a phos-acid, being that the synth presently existing for this that uses ethanolamine to reduce the iodide only gets 25% yield, this could possibly mean the possibility of achieving something more like 80%. alkylating amphetamine or PPA is definitely possible too.

loki · guinea pig

ah silly me i should read these things first... the problem with trying to get to a dimethylamine with methiodide is it is more inclined to go to the quat amine and plant 3 methyls on the amine, obviously not desirable.

i remember from the KrZ synthesis that it was found that doing the reaction in acidic conditions favoured the dimethylamine product. perhaps the same principle could apply here, the phosphate salt is probably a good choice. is the reason why this works to eliminate quat. formation because the acid occupies one of the potential bonds of the amine with the OH/ammonium hydroxide thingy?

if this is the case then using the DMT salt (whichever salt is most suitable) would thus prevent quaternary alkylation in the same way, except of course the addition of methiodide to dmt is a very rapid and complete reaction yielding around 90%, if it could be made to only produce the tertiary amine that would be wonderful.

back to the propargylation, the use of an acid to constrain alkylation would be helpful in the propargylation to narrow down the possible pathways too. i can see how if freebase was used it is likely it could form the N,N-dipropargyl, N-methyl amine.

here's a good question... what conditions are required for phosphite to perform the reduction? does it have to be acidic or can it be neutral or near neutral? i'm inclined to think that it can be done in neutral pH but i'm just speculating. my thought is that it simply needs to be acidic enough to ionise the amine and retain it in the aqueous solution where it is accessible to the phosphite. obviously one can't be doing any kind of useful reaction in alkaline conditions with an amine, in water.