The different isomers of 1-phenyl-2-nitroalcohol

[Bandil]

Hi!

Swim is working on a project with some different substituted 1-phenyl-2-nitroalcohol's, which are to be subjected to reduction and a certain cyclization process which gives some interesting compounds, and need some advice on the matter:

According to:

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

The (1R*,2S*) is formed in favorable amounts at low temperatures with a tertiary amine catalyst. This would yeild dl-norephedrine upon reducing. At high temperatures it yeilds the (1R*,2R*)/(1S*,2S*) version.

However, for the use that this compound is inteded the (1S*,2S*) could also be used(J Chem Soc, 850-854 (1952)). This isomer would yeild norisoephedrine upon reduction. Therefore it seems like overkill to monitor the temperature so carefully just to get one specific isomer, when both can be used.

The first article describes very throughly how the different isomers of the compounds are formed. It does not however deal with which isomers the old alkalihydroxide route yeilds. Especially

https://www.thevespiary.org/rhodium/Rhodium/chemistry/kamlet.phenyl-2-nitroalcohols.html

looks very interesting for general application for the synthesis of dl-nor(pseudo)ephedrine. But it doesn't really explain which isomers are mainly formed. Is it the (1R*,2R*)/(1S*,2S*) or the (1R*,2S*)/(1S*,2R*), that are maily formed here?

So i guess the real question is if the Kamlet method is as perfect as it looks, or does if form lots of the pseudo version?

Thank you for the aid

Regards
Bandil

[Rhodium]

Is it the (1R*,2R*)/(1S*,2S*) or the (1R*,2S*)/(1S*,2R*), that are maily formed here?

First of all when talking about stereoisomerism of formed products, the expression (1R*,2R*) means "mixture of (1R,2R) and (1S,2S)" and (1R*,2S*) means "mixture of (1R,2S) and (1S,2R)", thus writing "(1R*,2R*)/(1S*,2S*)" as you have done above is a bit redundant  

This now being resolved, on to your question:

In the experimental section of

European Patent 960,876

(https://www.thevespiary.org/rhodium/Rhodium/chemistry/phenyl-2-nitropropanol.html) you can find the sub-heading entitled Control III (Method of Kamlet), where they say that "HPLC analysis of the upper layer showed a conversion of 125.4 g. (69.3%) of total 2-nitro-1-phenyl-1-propanol with a (1R*,2S*)-isomer content of 43.9 g (35.1%)"

Thus, the net yield of (1R*,2S*)-2-nitro-1-phenyl-1-propanol using those conditions is 35.1% (based on benzaldehyde), and the remaining 34.2% thus consists of (1S*,2S*)-2-nitro-1-phenyl-1-propanol. Or, in other words, it seems like both diastereomers are formed in approximately equal amounts.

[Bandil]

Ahhhh, i see Thank you very much for clarifying that!

Another issue: I can't hep but wonder why the use triethylamine and not trimethylamine? It's not even mentioned in the patent. Maybe it gives so crappy yeilds that it is not even worth mentioning? It would seem like a little more obvious catalyst to at least try out...

Would that be worth a try?

Regards
Bandil

[Megatherium]

Trimethylamine seems to be a rather nasty substance (b.p.: 2.9 °C).

http://ptcl.chem.ox.ac.uk/MSDS/TR/trimethylamine.html

I think triethylamine is far more convenient to work with   .

[hest]

Yep, that's right Megatherium.

[Bandil]

Ok, i can see why that would be a bit nasty.

But there are some other things with the patent

Patent EP960876

that would be nice to tweak for the clandestine chemist:

They seem to be useing nitroethane and triethylamine as the solvent. They use a 4-5 times molar excess of each, and the triethylamine is supposed to be a catalyst(although in the table the use just under 1 molar eq. of the catalyst, which adds to the confusion). This isn't quite optimal, as it would be nice to conserve the nitroethane for better use(than solvent that is). Would it be feasible to make the same change as with the usual BA/nitroethane/NH4OAc route that Shulgin uses, into the one where an alcohol such as IPA is used as the solvent and the nitroethane in only used in equal molar amounts?

I would be nice to try out a system of 1:1:1 molar amounts of the benzaldehyde, nitroethane and triMETHYLamine(the one in 33% alcohol soln. which is commercial available) solvated in minimal amounts of cold IPA. Does any of you think that would be worth a shot or is there some obvious factor that i have missed that will make it fail miserably? I can't see why it's shouldn't work just as well... and i suppose there's only one way to find out!

Regards
bandil

[hest]

I'm sure that mixin one mole aldehyde with one mole nitroethane in cold alcohol with a sec. ammine as a cat will work. The amount of ammine is not important. Alternative a destilation of the product will give you all the nitroethane back.

[Antoncho]

Dear Hest - where d'that certainty come from? Do you know anything of sec. amines used as catalysts in Henry rxn?

SWIM once did exactly what you said above (unsubst'd BA) - and observed a seriously exotermic rxn even w/out heating, however, the post-rxn mixtr (a la Ba's MeAm condensation) refused to give xtals even after 1,5months of standing in the freezer at -20 C.

Nitroalcohol, i guess...

Do you know more than me about this?



A.

[Bandil]

Hest,Antoncho>

Isn't it tertiary amines that are used in this patent with the highest yeild?

Antoncho>
Swim has also observed the same with plain BA/nitroethane with an alkalihydroxide as catalyst. It simply made a yellow viscious liquid, which also refused to crystallize. There was never any attempt to purify it with DCM extraction / water wash, though. That might do the trick.

With the chance of getting some fire breathed down my neck my mr. Rhodium i would guess that the non-crystallized nitroalcohol(after water wash of course) is useable for reduction. The product can be purified via A/B afterwards.

Regards
Bandil

[hest]

Antoncho no only refs from the literature.
Beilstein gives 15 hits om nitroethane and benzaldehyde. the product is an oil so xtals is hard to get :-)
The synth is mixing nitroethane, benzaldehyde, a catalyst (amines K-t-buthanol silica ect.) and sometimes a solvent (THF) then let it react for 4-20 houers at 0-25°C.
Iff someone have a real interess I'll look into the literature (and doo some practical work) but I have a lot to doo, so if it's pure academic it has to wait
xxx
Hest

[Barium]

Primary and secondary amines can and will react with the aldehyde to give a imine. This imine will react with the nitroalkane to give a nitrostyrene. Tertiary amines, like triethylamine, can't form a imine at all. But the proton-ripping nature of them gives perfect conditions for the Henry-condensation to take place - forming nitroalcohols. Using a secondary amine one will get some nitroalcohol as a byproduct but the nitrostyrene will be the major product.

To isolate the nitroalcohol after a Henry-condensation using a alkali hydroxide I would suggest that you acidify the reaction mixture to liberate the nitroalcohol. Since a nitro group is weakly acidic it forms salts with bases.

1-Phenyl-2-nitropropene is notoriously difficult to crystallise. Make some seed crystals and add them to the oil and let the mixture stand. As Rhodium said, once you have some crystalline product save some of it because it will save you a lot of headache the next time.

[Rhodium]

Acidify slowly with dilute acetic acid when you want the nitro alcohol. Fast acidification with a strong acid (or inverse addition, dripping the reaction mixture into acid) catalyzes elimination of water, instead giving the nitroalkene as the main product.

[Barium]

Yes, I forgot to mention that. The sodium salt of a nitroalcohol added to a strong mineral acid, like hydrochloric acid, will give the nitrostyrene as the major product through elimination of one molecule water [1]. A weaker acid like acetic acid carefully added to the sodium salt of a nitroalcohol will not cause the elimination of water but simply liberate the nitroalcohol [2].

[1] Ar-CH(OH)-CH=NO2Na added to a excess of HCl --> Ar-CH=CH-NO2 + H2O

[2] HOAc carefully added to Ar-CH(OH)-CH=NO2Na --> Ar-CH(OH)-CH2-NO2 + NaOAc

[hest]

And bingo wee can make all the cat's and mar's of ouer favorite phenetylamines :-)
two book's Cihkal and Mihkal *lol*

[Bandil]

Swim tried out the reaction with triethylamine as catalyst in a methanolic medium, and used the molar eq. 1:1,1;1,1 of the subst. benzaldehyde, nitroethane, triethylamine. I will post the writeup once it has been purified, but i have some questions about the reaction conditions, which i find somewhat strange.

The patent

Patent EP960876

, says that "The reaction could, however, also be conducted in the presence of an inert reaction solvent. Common solvents, such as aliphatic alcohols, aliphatic and aromatic hydrocarbons, and others could be utilized.". Thus the solvent of choise was methanol. IPA might have been better, because of the lower polarity.

Originally, they use the nitroethane as a solvent and reactant. They also use VERY small amounts off it, considering that the reaction temperature is sub zero temperatures. How did they go about dissolving anything in this? Swim had to use 150 mL's of methanol to get some of the 7 g's of substituted benzaldehyde into the solution at the low temperatures. And at that dilution point only about 70% went into the solution. The reaction was carried out for two hours at sub zero temperatures, and was then allowed to return to 10 degrees above freezing very slowly. At about zero degrees everything had dissolved into the soln.

So what im pondering here, is it nescassry to use such HUGE amounts of solvents here, to get everything to dissolve, or is a slurry ok, as long as some of the benzaldehyde's dissolve, which then would react and perhaps allow some more to be dissolved?

Oh well, we'll see how it goes. I'll post details on the workup and reduction later today.

Regards
Bandil

[Barium]

How did they go about dissolving anything in this?

That patent is about making 1-phenyl-2-nitropropan-1-ol. Since they use benzaldehyde there is no problems getting a homogenous solution. With substituted benzaldehydes like the ones you want to use it's another story.

[Bandil]

So what would you suggest? Huge amounts of solvents, or maybe raise the temperature a little and get more of the norisoephedrine(which -should- be useable for the aminorex'es)?

[Barium]

Large amounts of solvents lowers the concentration of the reactants and will probably result in longer reaction time. Perhaps DCM could be used as solvent instead of the lower aliphatic alcohols. The substituted benzaldehydes should be quite soluble in DCM at lower temperatures (this is just a me guessing). Perhaps the preferred concentrations could be obtained then?