Author Topic: 4-Methylaminorex Synth w/o CNBr  (Read 17861 times)

0 Members and 1 Guest are viewing this topic.

psychokitty

  • Guest
Re: 4-Methylaminorex Synth w/o CNBr
« Reply #60 on: October 23, 2001, 05:22:00 AM »
:(

Rhodium

  • Guest
Re: 4-Methylaminorex Synth w/o CNBr
« Reply #61 on: October 23, 2001, 12:58:00 PM »
Not your nitroalcohol synth, but one using NaOH/HCl to give nitroalkenes. That is good, not bad.

halfapint

  • Guest
Re: 4-Methylaminorex Synth w/o CNBr
« Reply #62 on: November 02, 2001, 12:59:00 PM »
Re United States Patent  4183861 

The reaction temperature is maintained at 70.degree. C. and the pressure within the autoclave rises to a maximum of 2.4 atmospheres. The reaction is completed within three hours.


That seems to bee within the stability range for the SRV. Your two-liter polycarbonate autoclave.


turning science fact into <<science fiction>>

psychokitty

  • Guest
Re: 4-Methylaminorex Synth w/o CNBr
« Reply #63 on: November 04, 2001, 01:50:00 AM »
Okay, now to clear up a few things.

Rhodium:  I've been pretty confused lately because I couldn't understand why dimethylaminorex could be synthesized by BOTH the cyanate and cyanogen bromide routes.  Until, of course I figured out--by carefully examining the numbering scheme used--that 3,4-dimethylaminorex is (according to J. Med. Chem. Vol. 6, May, 1963, pp.266-272) 2-Imino-3,4-dimethyl-5-phenyloxazolidine and NOT 2-methylamino-4-methyl-5-phenyl-2-oxazoline (which has the basic 4-MAR structure with an N-methyl group at the 2-nitrogen). 

Of this I am certain.  The paper claims that using ephedrine in the standard cyanogen bromide cyclization reaction will yield an imine substance (namely this '3,4-dimethylaminorex') which does NOT belong to the aminorex family of compounds and which has the same mp as the 3,4-dimethylaminorex on you page (71-73).  This is very important because such a structural change negates most of the valued pharmacological effects of the aminorex class of drugs.

Just for informations sake, at the end of the article, the authors indicate that the appetite supressant pharmacology of each aminorex analogue varies considerably.  They also allude to the unique cardiovascular and CNS effects and claim that such effects would be the primary focus of an upcoming paper:

"Reference 17:  J. F. Gardocki and J. Yelnosky, manuscript in preperation."

I wonder if they ever wrote the damn thing.  A catalog of various aminorex analogues and thier corresponding CNS pharmacology would be most interesting and useful.

Anyway, the authors synthesized the "3,4-dimethylaminorex" by way of reacting cyanogen bromide and ephedrine.  They also formed another imino compound from the same reaction using phenylephrine as the starting material.  So, it is obvious that 2-alklamino aminorex compounds cannot be synthesized by reacting amino alcohols with cyanogen bromide.

However, the author indicated an alternative synthesis whereby an isocyanate was reacted with ephedrine which then formed an intermediate urea, which was then chlorinated with thionyl chloride and then cyclized by boiling in water. 

The original Poos U.S. Patent 3,161,650 indicates that although the intermediate ureas were best converted into the 2-alkylamino substances via thionyl chloride (via a displacement reaction) "other reactants susceptible to the type of reaction may also be used as, for example, hydrohalic acids such as hydrochloric acid, hydrobromic acid, or hydriodic acid . . .".

Sooooooo, it would appear that if the N-carbamyl-pseudoephedrine listed on the first page of this thread is indeed the intermediate urea Poos et al is talking about, then it should--with either thionyl chloride or hydrochloric acid (which, of course they use)--react through displacement and cyclization to form the expected 2-methylamino-aminorex. 

Again, however, that doesn't seem to be the case.  On the first page of this thread, Rhodium, N-carbamyl-pseudoephedrine is NOT converted to 2-methylamino-aminorex like one would think but instead to 3,4-dimethylaminorex.  Either you typed up the information incorrectly, or pseudoephedrine UNLIKE ephedrine forms the imino substance when used in the cyanate reaction instead of the 2-alkylamino compound.

BTW, according to the J. Med. Chem. paper potassium cyanate is effectively used to convert 2-amino-phenylethanol into the intermediate urea.  If my guess is correct, this can be cyclized into aminorex, making the cyanate reaction a good way to get to aminorex itself.

In the paper, the authors list former ways utilized to get to aminorex via various starting materials:

Reacting styrene oxide with sodium cyanamide led to a mixture of products from which a low yield of aminorex was isolated:  Ann. 467, 240 (1928).  Perhaps using a PTC in this reaction could help to raise the yields and overall efficiency of this reaction scheme.  Using stabilized cyanamide solutions and cyanamide itself under various cond

From styrene dibromide and urea:  (J. Pharm. Japan 449, 561 (1919).

For those curious about the numbering scheme used to determine the structure of the heterocyclic ring: one starts at the only oxygen ether linkage; two is the next amino group at the top point of the pentagonal structure(there is a double bond between number two amino group and number three amino group); three is the next amino group; four is the first methyl group; and five is the only methyl group directly attached to the benzene ring.

So the million dollar question is this:  Does N-carbamyl-pseudoephedrine cyclize with HCl to form 2-methylamino-aminorex or 3,4-dimethylaminorex?

--PK

PEYOTE

  • Guest
Mechanism
« Reply #64 on: November 04, 2001, 10:04:00 PM »
I've made a mechanism for the condensation of urea to 1-phenyl-1-hydroxy-2-aminopropane.

First: activation of urea



Second: attack of HNCO to norephedrine



Here you are. What about it?


JustATouchOfMojoHand

Rhodium

  • Guest
Re: Mechanism
« Reply #65 on: November 05, 2001, 02:39:00 AM »
Psychokitty: You are correct, the 3,4-dimethylaminorex I talk about is 4-MAR with a methyl group on the heterocyclic nitrogen (and thus the compound cannot form the two imine/amine resonance structures that aminorex and 4-MAR can do.

Do I have a 3,4-dimethylaminorex compound on my page? Where?

Correct, any 2-alkylamino aminorex compound cannot be formed either with CNBr or potassium cyanate, but it can through alkyl isocyanates, followed by cyclization of the intermediate urea (either with HX or with SOCl2).

I don't understand your confusion regarding the cyclization of N-carbamyl-ephedrine. This compound definitely cyclizes to 3,4-dimethylaminorex, and not to 2-methylamino-aminorex. Note that in the cyclization the urea nitrogen becomes the 2-amino group, and the ephedrine nitrogen ends up in the 3-position (with a methyl group attached).

Ephedrine can be converted with KOCN to 3,4-dimethylaminorex, while pseudoephedrine instead forms an amide (due to stereoconfigurational reasons), no 2-alkylamino compounds can be formed either with KOCN/CNBr.

Correct again, 2-amino-phenylethanol/KOCN should form an urea that can be cyclized into aminorex.

Could someone please find that old journal (J. Pharm Japan)? I would also appreciate any other articles regarding phenylpropanolamine derivatives which are reacted with KOCN to form ureas, followed by cyclization.

PEYOTE: What is your reference for that reaction? I did not know that urea could be transformed into HOCN like that?

Edited to fit the changed stereochemistry in the first post in this thread.

PEYOTE

  • Guest
Re: Mechanism
« Reply #66 on: November 05, 2001, 09:03:00 PM »
I havent got (for now) any ref for that rxn, because I've done it just this afternoon in laboratory:
I've made a phenylurea from an aniline derivative (phenetidin) with urea and HOAc,
and I can assure you that THAT'S correct (have you ever heard about synthons, or sinthetic equivalents?)
I'll serch for that refs,
but I've got a lot of things to do and very little time to do that.
Sorry for the 2nd part, I'ven't read it.... eh eh...



psychokitty

  • Guest
Re: Mechanism
« Reply #67 on: November 06, 2001, 05:56:00 AM »
Rhodium: Sorry for the confusion.  I went back and looked at the J. Med. Chem. paper and recognized where I made my mistake.  It would seem that all reactions involving ephedrine or pseudoephedrine would end up forming the 3,4-dimethylaminorex (which still really isn't from the aminorex family--at least Poos doesn't think so). 

I got mixed up thinking that the authors were reacting ephedrine with potassium cyanate, when in actuality, they were reacting norephedrine with methylisocyanate to form the 2-methylamino-aminorex.

The cyanate is regarded by the paper as "classical scheme" detailed in the following reference:

J.W. Cornforth, "Heterocyclic Compounds," vol. 5, R.C. Elderfield, Ed., John Wiley and Sons, Inc. New York, NY, 1957, p.384.

As for the phenylethanolamine, it doesn't seem to have any stereochemistry (maybe just dextro and levo).  So since it obviously can work in forming the intermediate urea, potassium or sodium cyanate should work at cyclizing it to aminorex.

As for phenylpropanolamine, according to the Merck index, it is norephedrine (I think).  Norpseudoephedrine, however, should be easily obtained from it through reaction with base or acid.  Don't know the exact ratios, but I do know there are more quantitative syntheses for pseudoephedrine from ephedrine than vice versa.

--PK

foxy2

  • Guest
Re: Mechanism
« Reply #68 on: November 06, 2001, 07:49:00 AM »
Aren't those naming schemes FUN to try to figure out!!
;D

They should always draw the fucking molecule IMHO

Do Your Part To Win The War

PEYOTE

  • Guest
Re: Mechanism
« Reply #69 on: November 06, 2001, 10:04:00 PM »
I've calculated the yield:

From 0.28 g of phenetidine hydrochloride (p-CH3CH2OC6H4NH3+Cl-) I've obtained 0.31g
of impure p-ethoxyphenylurea (theoric 0.29 g, so the yield is over 100%: 107%), this was recrystallyzed from water,
yielding 0.26g of recrystallized product (global yield: 89.65%). An IR spectra (KBr disc) confirms that
it's the desired product. Tomorrow I'll make an 1H-NMR.

Rhodium

  • Guest
Re: Mechanism
« Reply #70 on: November 06, 2001, 11:32:00 PM »
If it was recrystallized from water, perhaps your compound forms a hydrate of some sort? Now for the big question, is this reaction general for all amines, or only anilines?

If it is a general reaction, then we can make 4-MAR from phenylpropanolamine, urea, acetic acid and hydrochloric acid.

PEYOTE

  • Guest
Re: Mechanism
« Reply #71 on: November 07, 2001, 08:09:00 PM »
No, we used water because 4-ethoxyphenylurea isn't soluble in cold water but in hot water (this is general theory...) , I dont think that it forms any type of hydrate.
Well, I'm sure this is a general procedure, not only for aromatic amines but also for aliphatic amines. I'll search for some refs.

PEYOTE

  • Guest
Re: Mechanism
« Reply #72 on: November 09, 2001, 08:59:00 PM »
Here you are, from my laboratory notepad:

In a 50 mL round-bottom flask fitted with a condenser, pour 0.29g of phenetidine hydrochloride (1.67 mmol) and 2 mL of distillated water. Drop a solution of NaHCO3 10% until it reach pH 6.5 (about 4-5 drops). Then put 0.39 g of urea (6,49 mmol, 4 times of the equivalent of the phenetidine hydrochloride) and 4 drops of glacial acetic acid. Reflux for 1 hour; the reaction is monitored by TLC (eluant: CH2Cl/EtOH 19:1). The white precipitate is chilled, filtered (gooch) and dessiccated overnight on CaCl2. Yield: 0.31 (over 100%!!) So I make a recrystallization using water. Yield: 0.26 g of the recrystallized product, dulcine, or p-ethoxyphenylurea; total yield of the process: 93%. This is the reaction:

p-EtO-C6H4-NH3+Cl- + NH2CONH2 =(GAA)=> p-EtO-C6H4-NH-CO-NH2

1H-NMR analysis (with DMSO-d6;TMS 0,05%):

delta      proton type      signal/relative area

7.94      Ar-NH-CO-NH2   singlet/2
7.44 and 6.75   aromatic H      two doublet / 2 + 2
5.18      Ar-NH-CO-NH2   singlet/1
4.02      CH3-CH2-O-   quartet/2
1.38      CH-CH2-O-   triplet/3

PEYOTE

  • Guest
Further developments?
« Reply #73 on: November 29, 2001, 11:14:00 AM »
This document is archived with pictures at

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



Extracted from Chem. Reviews, 44, 447-476 (1949)




A. Pseudoureas (2-amino-2-oxazolines)

The pseudoureas, or 2-amino-2-oxazolines (XLI), are the most widely studied



of the substituted 2-oxazolines. Equilibrium with the 2-iminooxazolidine is possible, and the chemical properties indicate many reactions in both forms. The term ‘pseudourea’ arises from the fact that these compounds are isomers of alkenylureas; thus 2-amino-2-oxazoline is isomeric with N-vinylurea and hence is ‘ethylene pseudourea’; 2-amino-5-methyl-2-oxazoline is isomeric with allylurea and is known as ‘propylene pseudourea’. The compounds are solids and strong bases, forming well-characterized salts (58). 2-phenylamizlo-2-oxazolines have been suggested for use as local anesthetics (64), but other possible applications have not developed from rather extensive theoretical studies.

1. Syntheses of pseudoureas

(a) From beta-haloalkylureas: The hydrochloride of a pseudourea is obtained upon heating the haloalkylurea with water; the addition of alkali releases the free base.



The yields vary from moderate to quantitative. The necessary unsymmetrically substituted ureas are available via several routes. One is the method of Takeda (75), who prepared pseudoureas by heating styrene dibromide, or substituted styrene dibromides, with urea. Gabriel (38) prepared unsaturated ureas from unsaturated amines and isocyanates; the addition of a halogen acid to the double bond then gave the requisite beta-haloalkylurea. The product was cyclized to the pseudourea (XLII) on heating with water and to the imidazolidone (XLIII) with alcoholic potassium hydroxide.



The haloalkylureas can be prepared by the addition of iodine isocyanate to an olefin to form a beta-iodoisocyanate, which, when allowed to react with ammonia or an amine, yields the substituted urea. This method was studied by Birckenbach and Linhard (11, 12) in the synthesis of a series of cyclohexane derivatives.



These investigators also used olefins other than cyclohexane; however, the mode of addition of iodine isocyanate to unsymmetrical olefins is not known, and the position of groups in the pseudourea is then difficult to ascertain. Thus, they were able to convert s-phenylmethylethylene to the pseudourea in 94 per cent yield, but whether it had the structure XLIV or XLV was not determined; the compound was partially resolved, the d-form being isolated.



Another route to the beta-haloalkylureas is that originally used by Gabriel (32), in which a beta-haloamine is allowed to react with cyanic acid. In this instance, the pseudourea may be obtained directly, without isolation of the intermediate beta-halourea.
4-Keto derivatives of pseudoureas are formed by cyclization of alpha-haloacyl ureas, RCHBrCONHCONHR', on treatment with alkali. Examples of such reactions have been reported by Aspelund (2a). He has also reported the conversion of 1,5-diphenyl-5-bromobarbitudc acid to a pseudourea in a reaction which apparently involves formation of an alpha-haloacylurea as intermediate. Erlenmeyer and Kleiber (26a) report the formation of 5,5-diethyl-4-imino-4-oxazolidone from guanidine and ethyl alpha-hydroxy-alpha-ethylbutyrate.

(b) From beta-hydroxyalkylureas and thioureas: The preparation of a pseudourea from a beta-hydroxyalkylurea or thiourea requires a loss of water or hydrogen sulfide in the cyclization and is otherwise similar to the preparation from beta-haloalkylureas. In using this method Soderbaum (72) cyclized a beta-hydroxyalkylurea by heating with hydrochloric acid or by heating the thio analog with alcoholic mercuric oxide. The compounds prepared were derived from s-diphenyletahanolamine through conversion to the urea or thiourea with isocyanates or isothiocyanates. The reaction was carried out where R, in XLVa, was H, CH3, C2H5, C6H5, and o-CH3C6H4.





The reaction apparently takes a deferent course when -COOR replaces -CSNHR in XLVa. Thus, ethyl N-beta-hydroxyethylcarbamate, HOCH2CH2NHCOOC2H5, loses alcohol on heating to form an oxazolidone, not a 2-ethoxy-2-oxazoline (27a).

(c) From sodium cyanamide and chlorohydrins: Synthesis of pseudoureas (see table 4) from sodium cyanamide and chlorohydrins has no similarity to the syntheses of oxazolines as do the two preceding syntheses. Fromm and coworkers (28, 29, 30, 31) have described the reaction. The mechanism is not clear, but Fromm proposes that the sodium cyanamide reacts with the water present to form sodium hydroxide, which in turn dehydrohalogenates the chlorohydrin to the epoxide. Reaction of the epoxide (XLVI) with cyanamide then follows to form an unstable intermediate (XLVII), which rearranges to the pseudourea.



This mechanism is supported by the isolation of the pseudourea from the reaction when an epoxies is substituted for the chlorohydrin. Reaction between the chlorohydrin and sodium cyanamide to give the intermediate XLVII is not excluded by this evidence. The method has been successfully used with ethylene chlorohydrin and glycerol dichlorohydrin. A yield of 33 per cent of 2-amino-2-chloromethyl-2-oxazoline was reported from the epichlorohydrin. Reaction with chloroacetic acid gave an undescribed product.

References:

(2a) Aspelund, H. : Acta Acad. Aboensis, Math. et Phys. 12, No. 5, 33 pp. (1939) (Pub. 1940); Chem. Abstracts 41, 2413 (1947); 33, 6801 (1939); 35, 2143 (1941).
(11) Birckenbach, L,.and Linhard, M. : Ber. 64B, 961-8 (1931).
(12) Birckenbach, L,.and Linhard, M. : Ber. 64B, 1076-87 (1931).
(26a) Erlenmeyer, H., and Kleiber, A.: Helv. Chim. Acta 21, 111-12 (1938).
(27a) Franchimont, .A. P. N., and Lublin, A.: Rec. trav. chim. 21, 45-55 (1902).
(28) Fromm, E., Barrenscheen, H., Frieder, J., Pirk, L., and Kapeller, R.: Ann. 442, 130-49 (1925) .
(29) Fromm, E., Kapeller-Adler, R., Friedenthal, W., Stangel, L., Edlitz, J., Braumann, E., and Nussbaum, J.: Ann. 467, 240-74 (1928).
(30) Fromm, E., Kapeller, R., Pirk, L., Hahn, A., and Leipert, T.: Ann. 447, 259-84 (1926).
(31) Fromm, E., and Honold, E.: Ber. 55, 902-11 (1922).
(32) Gabriel, S.: Ber. 22, 1139-54 (1889).
(38) Gabriel; S., and Stelzner, R.: Ber. 28, 2929-38 (1895).
(58) Menne, E.: Ber. 33, 657-65 (1900).
(64) Rose, C. L., Schonle, H. A., and Chen, K. K.: Pharm. Arch. 11, 81-9 (1940); Chem. Abstracts 35, 1522 (1941) .
(72) Soderbaum, H. G.: Ber. 28, 1897-1903 (1895).
(73) Strauss, E.: Ber. 33, 2825-30 (1900).
(74) Takeda, J.: J. Pharm. Soc. Japan 426, 691-709 (1917); Chem. Abstracts 11, 3241 (1917).
(75) Takeda, J., and Kuroda, S.: J. Pharm. Soc. Japan 449, 561-608 (1919); Chem. Abstracts 14, 179 (1920).



scarmani

  • Guest
Re: Further developments?
« Reply #74 on: January 29, 2002, 08:41:00 PM »
I apologize for beating a dead horse.   :-[   I HAVE read Rhodium's several explanations and clarifications.   A couple times.  Most of my questions were answered but I was still somewhat unclear.   ::)




There are several precursors mentioned: ephedrine, pseudoephedrine, norephedrine, norpseudoephedrine and phenylethanolamine.  Different precursors will lead to different products.

Next, different precursors may each have varying stereoconfigurations depending on their source (OTC, synthesized, extracted, etc).   Again, different starting stereoconfigurations will lead to different products.

After that, (I gather that) any of these precursors can be used in one of two routes: the cyanogen bromide route and the cyanate route.  Once again, the choice of route can affect the product produced.

Finally there are several possible mentioned products: aminorex, 4-methylaminorex, 3,4-dimethylaminorex, and 4-methyl-5-phenyl-oxazolid-2-one.  Each compound is of different pharmacological properties.

And, of course, the active products have different stereoconfigurations which are also crucial to their pharmacology. There is the optimal stereoconfiguration of 4-MAR, then the other configuration that's five times less potent.  Then of course there is the lovely isomer of (x,y?)dimethylaminorex that causes fatal seizures.

Postscript: it is not entirely clear exactly what sources of precursors have which stereoconfigurations.




Given all these branches I am not overly ashamed that I can't get a complete grasp of the TREE, starting with PRECURSOR + SOURCE and ending with PRODUCT + PHARMACOLOGY.

Can anyone diagram or outline the "before and after" in a single post?  That would bee cool  8)

Those innocent eyes slit my soul up like a razor.

SPISSHAK

  • Guest
About those isomers?
« Reply #75 on: April 18, 2002, 07:30:00 AM »
I was just curios since only the Levo-isomer of the amino alcohol yeilds the desired 2-amino-oxazoline, is this the same isomer which is the phenylpropanolamine of commerce?

Another thing, about the mechanism of the forementioned cyanate route.
As far as I can tell, the cyanic acid adds directly to the amine and rearranges to give the substituted urea intermediate which is isolated by recrystallization; then, the urea derivative with the alcohol grouping adjacent to it is refluxed in HCl forming the halide which then undergoes elimination of H-X by abstracting a proton from the urea group, then, this forms an Imine (and a carbocation??) and the carbcation bonds to the oxygen of the urea group forming the ring , the Imine component then abstracts a proton from the former amine component of the amino alcohol to form the oxazoline ring with the amine in the two position. Correct me if I`m wrong or just flame away.
But back to the original question about the mechanism involving stereochemistry in the cyanate route.
How does amide formation take place with the dextro isomer of the phenylpropanolamine Is this due to the relationship of the alcohol grouping to the amine?
for example with either the -,- isomer or, the
+,+ isomer the alcohol grouping is (cis so to speak) to the amino group and vice versa does this have something to do with it.
Comments and insight into these mechanisms would be greatly appreciated.

 
[/b] A mind is a terrible thing to educate by yourself. 8)  ;)

element109

  • Guest
look up the original article.
« Reply #76 on: April 19, 2002, 07:50:00 PM »
Ephedrine gives the aminorex analog, while pseudoephedrine will yield the amide. So *norephedrine* would yield 4-methylaminorex while norpseudoephedrine (also known as the d-phenylpropanolamine of commerce) yields the amide.

It's the position of the OH-group that matters here.
That's why we should all try to dig up the Russian ref given by Psychokitty in the "Wanted references" thread in the novel discourse, which details the isomerization of pseudoephedrine to ephedrine.

I have the article and I'm willing to outline both reaction mechanisms.

e109

SPISSHAK

  • Guest
isomerization of amino acohols
« Reply #77 on: April 19, 2002, 11:08:00 PM »
So when you speak of isomerization of the pseudoephedrine to the ephedrine, you are refering to the alcohol groupsin relation to the amine.
and not whether the amine is D or,L.
If this is the case thaen that can be done wih 25% hydrochloric acid.
one caveat to this procedure is that it works by converting the alcohol to the chloride and subsequent hydrolysis in a reversible reaction, which gives the isomer of the alcohol grouping.
The paper published by Emde (on rhodium`s website) is wrong with respect to the temperature at which this occurs at the temperature given (80 degrees celcius), the hydrochloric acid will dehydrate the alcohol and amine to give an aziridine which is uterly useless for our purposes.
This is evidenced by a darkening of the solution of ephedrine and hydrochloric acid.
I would say that isomerization of the alcohol grouping should be done at 40 degrees centigrate, and prolong the reaction time, also maybe play with the acid concentration a little, if that does`nt work.

SPISSHAK

  • Guest
about that mechanism?
« Reply #78 on: April 19, 2002, 11:23:00 PM »
I would be grateful if you could outline those mechanisms.

SPISSHAK

  • Guest
R,S,S,R isomers. isolation of R,R, or S,S isomers
« Reply #79 on: April 20, 2002, 10:44:00 PM »
Another good idea which may be useful to our ends is the isolation of R,R isomer via preciitation of the sufuric acid ester.
I`ll dig up the article in my next post but essentially as rhodium stated in his  first post, that either R,R or S,S isomer will work to for the aminorex via the cyanate route.
but, the R,S or S,R isomer will give an amide.
So if the phenylpropanolamine used here is racemic, then maybe this reaction to form the sulfuric ester of ephedrine
could be used in this case with phenylpropanolamine.
but in the example I will give in the next post R,S and R,R isomer of psuedoephedrine is added to chilled sulfuric acid
then this is precipitated by adding crushed ice and alcohol.
only the R,R isomer precipitated.
Which is the isomer of interest in this scheme.
Then, I suppose it could be hydrolysed by whatever means to give back the starting ephedrine, I will put this in the next post after I locate it, referenced and all.