Author Topic: Norpseudoephedrine synthesis & optical resolution  (Read 5210 times)

0 Members and 1 Guest are viewing this topic.

Rhodium

  • Guest
Norpseudoephedrine synthesis & optical resolution
« on: June 06, 2003, 03:20:00 PM »
Stereospecific synthesis, and optical resolution of 2-amino-1-phenylpropanol (Norpseudoephedrine)

Patent US5962737



Abstract

Stereospecific synthesis of the racemic threo isomers of 2-nitro-1-phenylpropanols by reacting a benzaldehyde derivative with nitroalkane in the presence of a tertiary amine and reducing 2-nitro-1-phenylpropanols with, for example, lithium aluminum hydride to 2-amino-1-phenylpropanols is described. Also described are phase transfer resolution of racemic mixtures of 2-amino-1-phenylpropanol and its derivatives into their optically pure isomers by reacting a racemic mixture with the mono alkali metal salt of a tartaric acid ester in a two phase system of a hydrocarbon and water. The specification further describes therapeutically useful optically pure isomers of threo-2-amino-1-(dialkoxy or alkoxy) phenylpropanols and acid addition salts thereof.


Prior Art

Mixed stereoisomers of 2-amino-1-phenylpropanol can be readily prepared by reacting benzaldehyde with nitroethane in the presence of an alkaline catalyst to produce 2-nitro-1-phenylpropanol which is then reduced to the amine. E.g., Hoover and Hass, Journal of Organic Chemistry, 12, 506, (1947). This reaction gives excellent yields at low cost. However, separating the stereoisomers produced by this reaction has not been satisfactory. As a result, heretofore, the only practical stereospecific synthesis of 2-amino-1-phenylpropanols involves the reduction of propiophenone derivatives to the racemic erythro diastereoisomers and fractional crystallization of the amine salt of an optically active acid.

The DL-erythro forms can be readily obtained from the appropriate propiophenone derivatives or by inversion of DL-threo derivatives. Since the pharmacological properties of the D and L isomers differ, it is desirable to separate the two, resulting in maximum therapeutic utility. While such potentially useful products can also be prepared, the difficulty of separating, i.e. "resolution", the isomers has prevented their development.

Specifically, the base is converted to a salt of an optically active acid. For example, the DL-base is reacted with a D-acid. This results in the formation of a mixture of D-base-D-acid and L-base-D-acid. These two salts differ in solubility. When the mixture is cooled and allowed to stand, the D-base-D-acid, typically being the less soluble, precipitates out of the solution first. By removing the precipitate at the appropriate time, the collected precipitate is largely the D-base-D-acid, while the L-base-D-acid remains in solution.

While this results in purification of the optical isomers to some extent, substantial impurities often remain. These impurities are removed to some degree by fractional crystallization. The "pure" D and L bases are then liberated by adding a sufficient amount of an alkali to a solution of the "pure" salt to produce a pH above 7.

The above process is very tedious, time consuming, and inefficient. In addition, a base with the same configuration as the resolving acid is more easily separated and purified than the opposite enantiomer. Consequently, in practice, the desired isomeric base must be matched with an optically active acid of the same configuration to achieve maximum yields.


Description Of Invention

It has been discovered that if a benzaldehyde derivative is reacted with a nitroalkane in the presence of a tertiary amine, the nitroalcohol formed is of a threo configuration rather than a mixture of threo and erythro isomers which results when sodium hydroxide is used as the catalyst. The reaction is best conducted in an aqueous aliphatic alcohol.

The tertiary amine employed must be free of primary and secondary amines which interfere with the reaction. These can be removed from the commercial products by refluxing with acetic or phthalic anhydride followed by distillation.

In a preferred embodiment of the synthetic method of the present invention, a benzaldehyde derivative is reacted with nitroethane in the presence of triethylamine in aqueous ethanol. The mixture is allowed to react at room temperature for twenty-four hours. The mixture is then acidified with an organic acid since mineral acids tend to promote decomposition, as does heat. Excess solvents and reactants are evaporated and the nitroalcohol extracted.

The nitroalcohol can easily be reduced by catalytic hydrogenation or conventional reducing agents such as zinc and acid. Lithium aluminum hydride has been found to work exceptionally well. The racemic threo isomers can readily be inverted to the racemic erythro isomers by reaction with acetic anhydride and thionyl chloride and hydrolysis. N-methyl derivatives can be prepared by adding an equimolecular amount of aqueous formaldehyde to the primary amine and reducing the Schiff base The use of a primary aliphatic alkylamine results in the formation of the dehydration product of the nitroalcohol--the beta-nitrostyrene. By employing a secondary or tertiary amine in the presence of water, nitrostyrene formation is completely eliminated. The use of a secondary or tertiary amine prevents the formation of a Schiff base amine, the necessary intermediary for nitrostyrene formation.

Instead of the usual fractional crystallization of the amine-acid salt, the method of resolving stereo isomers of the present invention uses a novel phase transfer resolution. It has been discovered that if the amine base is rapidly stirred with a mono alkali metal salt of a tartaric acid ester in a two-phase system of a hydrocarbon and water, a rapid and efficient resolution can be achieved.

In a practical embodiment of the resolution method of the present invention, a DL-2-amino-1-phenylpropanol derivative in dichloromethane is combined and stirred with dibenzoltartaric acid in water, and aqueous sodium hydroxide for from about one to three hours. The reaction mixture is then allowed to stand for the about same length of time. The dichloromethane phase is separated and dried over anhydrous magnesium sulfate. Evaporation gives the L-threo isomer in nearly quantitative yield. The aqueous phase is made alkaline with ammonia and extracted with dichloromethane. The dichloromethane extract is dried over anhydrous magnesium sulfate and evaporated to give the D-threo isomer in nearly quantitative yield.


Example 1



A racemic mixture of threo nitroalcohols was prepared by combining freshly distilled benzaldehyde (1 mole), nitroethane (2.5 moles), and triethylamine (0.05 mole) in ethanol (150 ml.) with water (75 ml.). This mixture was allowed to stand at room temperature in the dark for twenty-four hours. The mixture was then ice-cooled and acetic acid (0.05 mole) was added to the reaction mixture. Alcohol and excess nitroethane were evaporated (vacuum). Water (75 ml.) was added and the nitro alcohol extracted with ethyl acetate, dried over anhydrous sodium sulfate and the solvent evaporated (vacuum) to give the product, a viscous oil (yield 70-80% based on the benzaldehyde).

Example 2



The nitro alcohols were reduced by two methods, a zinc and acid method and a lithium aluminum hydride method as described below:

(A) Zinc and Acid

Hydrochloric acid (4 moles) is added (with stirring) to a mixture of nitroalcohol (1 mole), zinc dust (4 moles), and 400 ml. of 95% ethanol. The acid is added at such a rate that the temperature remains at 45 degrees or below (several hours are usually required). Stirring is continued for 1-2 hours after completing the addition. The acid solution is extracted with ether to remove non-basic materials. Excess NaOH solution is then added and the free base extracted with ether. The ether solution is dried (MgSO4) evaporated, and the product distilled or crystallized in the usual manner (70-80% yield).

(B) Lithium Aluminum Hydride ("LAH")

A solution of the nitroalcohol (1 mole) in tetrahydrofuran (400 ml.) is added to a solution of lithium aluminum hydride (4 moles) in tetrahydrofuran (500 ml.) with rapid stirring and cooling (as necessary) to maintain gentle reflux. After the addition is complete, the mixture is refluxed for an additional 2 hours.

Water is added to neutralize excess LAH (1 liter) and the product extracted with benzene. The benzene extract is dried (MgSO.sub.4) and the benzene evaporated. The resulting product is purified by either distillation or crystallization (80-90% yield).

Example 3

The reaction mixture of reduced nitro alcohols was resolved into optically pure isomers by the following process.

A mixture of a DL-threo-2-amino-1-phenylpropanol (1 mole) in dichloromethane (600 ml.), dibenzoyltartaric acid (0.5 mole) in distilled water (30 ml.), and sodium hydroxide (0.5 mole) in distilled water (50 ml.) is stirred rapidly for two hours and allowed to stand for two hours. The dichloromethane phase is separated using a separating funnel over anhydrous magnesium sulfate. Rotary evaporation of the dichloromethane phase gives the L-threo isomer in nearly quantitative yield.

The aqueous phase is made alkaline with ammonia to pH 13 and extracted with dichloromethane. The dichloromethane extract is dried over anhydrous magnesium sulfate and evaporated to give the D-threo isomer in nearly quantitative yield. The enantiomeric purity of the products is 96-99% based on GLC analysis of the D or L--methoxy--trifluromethylphenylacetamide (MTPA) derivatives.

Bandil

  • Guest
The amine catalyst gives the nitroalcohol.
« Reply #1 on: June 07, 2003, 02:06:00 AM »
The amine catalyst gives the nitroalcohol. The invention suggest the use of an alkalimetal catalyst, yet the examples uses a tertiary amine.

How come the tertiary amine gives the nitroalcohol? Is it the fact that it cant for imines, which in turn reacts with the nitroalkane? Thats the only reason that i can think of ...

It's also quite a large Nitroalkane:BA ratio they use. Anyone care to elaborate on that?

Bandil


Barium

  • Guest
The invention suggest the use of an ...
« Reply #2 on: June 08, 2003, 06:56:00 AM »
The invention suggest the use of an alkalimetal catalyst, yet the examples uses a tertiary amine
No, the invention is about using a tertiary amine. The old method (prior art) used NaOH or NaOMe but gave lower yields due to problems in the work up.

How come the tertiary amine gives the nitroalcohol? Is it the fact that it cant for imines, which in turn reacts with the nitroalkane?
Exactly! Primary and secondary amines forms imines with the aldehyde which gives nitrostyrenes with nitroalkanes. Tert-amines can't form imines.

It's also quite a large Nitroalkane:BA ratio they use. Anyone care to elaborate on that? Nitroethane is a cheap-as-shit reagent for them, and thus used in excess to push the reaction towards maximum yield of the nitroalcohol.

What I'd like to see is them give a exapmple of a procedure where the nitroalcohol is reduced to aminoalcohol using low-pressure catalytic hydrogenation since it's soo easy according to them.  :P


Vibrating_Lights

  • Guest
AA
« Reply #3 on: June 08, 2003, 08:01:00 AM »
Is the Acetic acid, added after the condensation, Just there to make the Amine a salt so it is not extracted with the nitroalcohol?

Can the nitroalcohol be reduced with NaBH4/THf then Pd/BaSO4.

The NOR just means it is dl-ephedrine unmethylated correct. 

After reducing the nitroalcohol to the amphetamine could it be used as is in the KOCN PPA route to 4 MAr?


Rhodium

  • Guest
Corrections for V_L
« Reply #4 on: June 08, 2003, 09:50:00 AM »
The acetic acid is added to liberate the free nitroalcohol, which before the addition was present as a salt with the amine. Add the acid slowly in dilute form as to minimize the formation of any nitroalkene.

Nitro groups are not reduced with sodium borohydride.

The "nor" prefix in norpseudoephedrine (not norephedrine, see the title) designates that there is no N-methyl group.

The product of this procedure can not be used in the KOCN 4-MAR procedure, as it consists of norpseudoephedrine rather than norephedrine.

Vibrating_Lights

  • Guest
???
« Reply #5 on: June 08, 2003, 10:09:00 AM »

Bandil

  • Guest
How come it's the not the racemic r-s form...
« Reply #6 on: June 08, 2003, 02:16:00 PM »
How come it's the not the racemic r-s form thats formed at the -OH bond? The carbonyl carbon on the BA is plannar, and thus the attack from the alpha-carbon on the nitroalkane should be able to happen from either side.

Do you have an explanation and/or reference on why it's only the pseudo thats formed, and not the racemic form?

Regards
Bandil


Vibrating_Lights

  • Guest
Can PPa
« Reply #7 on: June 08, 2003, 02:51:00 PM »
Can PPA not be made from benzyaldehyde???


Rhodium

  • Guest
Nitroalcohol stereochemistry
« Reply #8 on: June 08, 2003, 05:01:00 PM »
Bandil: The racemic threo-nitroalcohol precursor to norpseudoephedrine is actually formed (the (R,R) and (S,S) isomers), the enantiomers are separated first in Example 3. What this tertiary amine catalyst avoids is the formation of the (R,S) and (S,R) isomers (alias erythro) of 1-Phenyl-2-Nitropropan-1-ol (precursors to norephedrine). See the illustrations at

https://www.thevespiary.org/rhodium/Rhodium/chemistry/norpseudoephedrine.html

- it may help.

Vibrating_Lights: Norpseudoephedrine is racemic threo-Phenylpropanolamine (PPA). Are you thinking of Norephedrine, or erythro-Phenylpropanolamine? It can be made as a mixture together with Norpseudoephedrine if you use sodium hydroxide instead of triethylamine in the condensation step above, see the patent in its entirety:

Patent US5962737

I am not aware of any simple procedure for producing pure Norephedrine/erythro-Phenylpropanolamine alone, starting from benzaldehyde.

The patent also tells about the production of 3,4-Dimethoxy-, 2,5-Dimethoxy-, 3,4,5-Trimethoxy- and 3,4-Methylenedioxy-Norpseudoephedrine (Example 5-8), but not in much detail.

HumbleBee

  • Guest
Norephedrine from benzaldehyde
« Reply #9 on: September 11, 2003, 04:42:00 AM »
The DL-erythro forms can be readily obtained from the appropriate propiophenone derivatives or by inversion of DL-threo derivatives.

How is the procedure for inverting the DL-threo derivatives?

However there is a method for making PPA from benzaldehyde mainly resulting in norephedrine - excerpt from Patent

Patent EP0960876

:


Example 1
Nitroethane (10.2 g., 0.132 mole) was mixed with triethylamine (17.1 g., 0.169 mole), cooled to a temperature of -8 DEG C and benzaldehyde (5.1 g., 0.047 mole) added. After 2.7 hours at -10 DEG C, the mixture was neutralized. HPLC analysis showed a conversion of 8.25 g. (96.9%) of total 2-nitro-1-phenyl-1-propanol. 6.40 g of the 2-nitro-1-phenyl-1-propanol was the (1R*,2S*)-stereoisomer (77.6%).





The (1R*,2S*)-stereoisomer will give norephedrine upon reduction. Same reactands but different conditions...


Rhodium

  • Guest
Good post, I have now HTMLized the entire...
« Reply #10 on: September 11, 2003, 12:56:00 PM »
Good post, I have now HTMLized the entire patent and uploaded it here:

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



And also these, which are referenced from the above patent:

Amino Alcohols. I. Phenylpropanolamine and Para-Tolylpropanolamine
Hartung & Munch

JACS 51, 2264 (1929)

(https://www.thevespiary.org/rhodium/Rhodium/pdf/hartung.aminoalcohols-01.pdf)

Preparation of Arylnitroalkanols
Jonas Kamlet

US Pat. 2,151,517

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

Synthesis of 2-Amino-1-Phenyl-1-Propanol and its Methylated Derivatives
F. W. Hoover & H. B. Hass

J. Org. Chem. 506-509 (1947)

(https://www.thevespiary.org/rhodium/Rhodium/chemistry/nor-pseudo-ephedrine.html)

Organikum

  • Guest
norephedrine without norpseudo
« Reply #11 on: June 25, 2004, 04:59:00 PM »
Rhodium wrote:

I am not aware of any simple procedure for producing pure Norephedrine/erythro-Phenylpropanolamine alone, starting from benzaldehyde


I am.  :)
Benzaldehyde + acid hardened yeast/molasses/nutrient salts/thiamine = l-PAC (1)
l-PAC + NH3 = l-norephedrine (2)
(the raw etherreal extract from the biosynthesis can be used this a Al/Hg red. amination).
Overall yields for benzaldehyde to l-norephedrine are about 27% without optimization.

I am sure that it is possible to reach about 50% what would be absolutely satisfying IMHO as there are no hard to get reagents needed. Benzaldehyde is the only one and its synthesis is well covered. For the synthesis produces exclusivly norephedrine this might be worth to be considered if 4-MAR by KOCN is wanted.

(1) this is a variation of the known l-PAC biosynthesis without acetaldehyde.

Patent DD51651

First example of prior art and second example of the patent itself. Be aware that it is NOT possible to mix the procedures just as it comes to mind. In most cases this will LOWER yields. The named combination of acid-resistant yeast and thiamine works well though.
(2) hydroxyphenylacetones form quite stable imines unlike ketones. Therefor NH3 or a salt of NH3 can be used without risking bad yields.


French patent no. 844,228: "PROCESS FOR THE PREPARATION OF AMINES"
It's already known that reducing hydroxyketones or polycarbonylic compounds
with activated aluminum and ammmonia derivates leads to formation of the
corresponding aminoalcohols. This reaction is favored because of the fact that
hydroxyketones and polycarbonyls are apt to form relatively stable imines.



from:

https://www.thevespiary.org/rhodium/Rhodium/chemistry/alhg-pressure.txt




ORG  ;)




Vitus_Verdegast

  • Guest
why not oxime?
« Reply #12 on: June 25, 2004, 05:50:00 PM »

Benzaldehyde + acid hardened yeast/molasses/nutrient salts/thiamine = l-PAC (1)
l-PAC + NH3 = l-norephedrine (2)
(the raw etherreal extract from the biosynthesis can be used this a Al/Hg red. amination).
Overall yields for benzaldehyde to l-norephedrine are about 27% without optimization.


Why not make the oxime first and perform an acidic Al/Hg à la Antibody2 on that, yields should be moderate to high using that route, and it has the advantage that the oxime can be purified first, so the final amine will need less work-up to get a pure product.

Hydroxylamine salts are not hard to make, and they are used in photographic processes so not hard to get either.

(edit) Don't get me wrong, it is great news that NH3 can so easily be used here. I'm just posing a suggestion to improve yield and purity.


Organikum

  • Guest
Hydroxylamine quite sure would give better...
« Reply #13 on: June 25, 2004, 07:26:00 PM »
Hydroxylamine quite sure would give better yields. The argument regarding the workup does not convince me though. Where is the advantage? Actually you have to do two workups instead of one and a workup on a amine is the most easy one IMHO. A/B - B/A/B - B/A, recrystallization and thats it. Procedures taken from alkaloid extractions make life easy (like filtering the acidified solution through a filter with sodium carbonate etc.).

Not that hydroxylamine is not to come someday. As acetaldehyde will. Now the small scale of the reaction causes no into volumetric problems in the biosynth part and making some more benzaldehyde is a small problem compared to making additionally a complete other compound. The Al/Hg was known to work on l-PAC. Its hard enough to determine the points of failure in known reactions, at least for me with TLC as most sophisticated tool of analysis  ;D .

It was a preview on some work in progress anyways.
ORG


welt

  • Guest
Zn and H2SO4 ?
« Reply #14 on: June 25, 2004, 08:54:00 PM »
in the thread "4-Methylaminorex Synth w/o CNBr" by Rhodium
psycokitty says that the nitroalcohol can be reduced with Zn and aqueous H2SO4 (the post was labled "No 225556") does any one know if this true becasue it would be wonderful if it was

Vitus_Verdegast

  • Guest
org:
« Reply #15 on: June 26, 2004, 05:06:00 PM »
The argument regarding the workup does not convince me though. Where is the advantage? Actually you have to do two workups instead of one and a workup on a amine is the most easy one IMHO.

True. I was thinking along the lines of trying to obtain a solid oxime from the impure l-PAC, and purifying that by recrystallization, so there is a cleaner product available for the following reduction. But of course this will require more work and in many other cases, attempts to make oximes from undistilled P2Ps always resulted in oily oximes. So it wasn't really a good suggestion.


BOS

  • Guest
Some patents
« Reply #16 on: June 26, 2004, 09:14:00 PM »
These may have been posted already,sorry bout` that.
Some may find the reading fun.

Patent EP1142864



An efficient process for stereoselectively producing L-erythro-(1R,2S)-2-amino-1-phenylpropan-1-ol from L-(R)-phenylacetylcarbinol, which comprises reductively aminating L-(R)-phenylacetylcarbinol with a primary aralkylamine under catalytic reduction conditions and successively subjecting the resultant L-erythro-(1R,2S)-2-(N-aralkylamino)-1-phenylpropan-1-ol to catalytic reduction to remove the N-aralkyl group in a manner as in hydrogenolysis.

Patent GB365535



I-Phenyl-2-aminoalcohols-(1); oximes.--l-1-Phenyl-2-aminopropanols-(1) are prepared by (1) treating l-1-phenyl-2-ketopropanol-(1) with hydrogen and either (a) a precious metal catalyst in presence of ammonia or a primary or secondary amine excepting methylamine, or (b) a catalyst comprising iron, cobalt, nickel or copper in the presence of an ammonium salt or a salt of a primary or secondary amine; (2) converting l-1-phenyl-2-ketopropanol-(1) into its oxime with hydroxylamine and catalytically reducing it with a precious metal catalyst. The product of (2) may be alkylated to yield the corresponding alkylamino compound. Examples are given of the preparation of (1) l-1-phenyl-2-aminopropanol-(1) by treating l-phenylacetylcarbinol with hydroxylamine and hydrogenating the resulting oxime in acetic acid solution using palladium as catalyst, and (2) l-1-phenyl-2-methylaminopropanol-(1) by the hydrogenation of a solution of l-phenylacetylcarbinol and methylamine hydrochloride in alcohol in the presence of nickel. Specification 313,617 is referred to. The Provisional Specification describes also the conversion of optically active 1-phenyl-2-ketoalcohols in general into the corresponding 1-phenyl-2-aminoalcohols-(1) by the foregoing processes, and includes an example of the hydrogenation of l-phenylacetylcarbinol in alcohol solution in presence of methylamine using palladium as catalyst to form l-phenylpropanolmethylamine.

Patent GB365541



1 - Phenyl-2-aminoalcohols - (1).--Racemic 1-phenyl-2-aminopropanols-(1) are prepared by treating l-1-phenyl-2-ketopropanol-(1) with hydrogen in presence of ammonia or a primary or secondary amine using iron, nickel, cobalt, or copper as the catalyst. An example is given of the conversion of l-phenylacetylcarbinol into racemic 1-phenyl-2-methylaminopropanol-(1) by hydrogenation in the presence of methylamine and nickel. Specification 313,617, [Class 2 (iii), Dyes &c.], is referred to. The Provisional Specification describes also the conversion of optically active 1-phenyl-2-ketoalcohols-(1) in general into the corresponding 1-phenyl-2-aminoalcohols-(1) in the racemic form by the foregoing process.

Patent US4224246



A process for the synthesis and separation of the threo and erythro isomers of 2-amino-1-phenyl-1-propanol comprising the steps of catalytically reducing 2-nitro-1-phenyl-1-propanol to form the acetate salt of the racemic mixture of 2-amino-1-phenyl-1-propanol and separating the isomers by fractional crystallization.

Organikum

  • Guest
phenylacetylcarbinol is not so stable, ...
« Reply #17 on: June 27, 2004, 03:43:00 AM »
phenylacetylcarbinol is not so stable, purification and isolation are painful and bring serious losses. Elevated temperatures and basic conditions lead to razemisation and decomposition rather fast. This was already mentioned by Neuberg/Hirsch 1921 who discovered the compound. Only the lateron discovery of Hildebrandt/Klavehn that the raw extract of the yeast mediated transformation can be used in a reductive Al/Hg amination/alkylation made the process feasible for industrial use. (I am sure nowadays industry has found ways to do this separations and Al/Hg is - at least in the western world - not used anymore).
Of course as a rule a purified precursor mostly gives more and better product. Not here though. A double A/B workup on the final product is necessary and easy too. I guess all amines would profit from such a workup.