@German, there are references for Festers ester reflux. In the patent below example 1 is what you are talking about and comparative example 1 is the reflux but they use a 4 hour reflux instead of Festers 3 hour. The only difference is an increase in yields of 13%+ if done without reflux (adding extra methanol as it distills off).

Method for preparing amino acid esters
EP0544205B1
http://sharebee.com/15e06c76

".......it's an old topic of mine"

Yes I know, I have followed from way back at the Hive, to SM to ChemicalForums to various blogs(to name a couple ). Thanks for clarifying also, but I have to say PLEASE read the patent below carefully!

Process for preparing 1-phenyl-2-aminopropane
http://sharebee.com/8bcadfd2

They use 1-phenyl-2-aminopropanediol-1,3 which has the following constitution;

Using (16.7 grams = 0.1 mole) of 1-phenyl-2-aminopropanediol and (140ml = 1.1 mole) of 58.4% Hydroiodic acid and (6.2 grams = 0.2 atomic wieght)red phosphorous, refluxed at 123-125*C for 24 hours yielded 85.4% 1-phenyl-2-amino-iodopropane with trace amounts of 1-phenyl-2-aminopropane and even smaller amounts of 1-pheny-2-aminopropanol-3.
And there are other examples in that patent that use much more red p and HI!


They actually state this in the patent;

"As to the step for the reductive deiodonation of the compound (1-phenyl-2-amino-3-iodopropane), the reaction is carried out preferably at high temperature to produce the desired compound, 1-phenyl-2-aminopropane. Suitable reductive dehalogenating agents are Raney nickel and lithium aluminium hydride. Although Hydroiodic acid itself maybe effective as the reductive dehalogenating agent, in this case the reaction is too slow to be practical"

And they said that after pushing their reactions out to 35 hours!!! I have refluxed in excess of 48 hours, and adding another 48 would not be the end of the world. But just be aware if you want the complete reduction (dehalogenation) you are going to have to go stronger than 24 hours. Or at least do some serious experimenting like Java stated. Now I know you cant depend solely on one patent, and trust me I can produce more concerning reduction of primary alcohols with HydroIodic acid.

Further studies on the reduction of benzylic acids by hypophosphorous acd/Iodine
http://www.arkat-usa.org/get-file/18759/

The only primary alcohol reduced in the paper above was Cinnamyl alcohol, and except for dimerization the major product was the corresponding iodopropane in shitty 21% yields.


"the material was tested and no OH group showed up in the final analysis and on TLC the run matched a control sample....."

I dont expect their would be an OH group, as stated about 4-5 hours into reflux all the OH's will be displaced with I's...  I also dont imagine TLC would provide a clear reading here either, the retention time of the iodo derivative would be too close to descriminate (double bubble). Depending on what combination of solvents you used as your eluent, would determine how clear the resolution was. But that being said if it was N-methylated and on testing kept you up for 2 days, there is no disputing that. I have been talking about this alot, because I am somewhat concerned that people will take this easier route and will end up taking iodopropanes instead of the good stuff. I sincerely appreciate your input Java, and was hoping to grill you further on Aziridines in this thread.

The 2nd step confused me. How do you intend to alkylate the tertiary carbon? Can't think of a way to realize this. So maybe it was indeed crazy

99,99999% that it won't work, but here is an idea:



The phenylalanine is converted with the Hunsdiecker reaction to an intermedier what is unstable as hell, this would be an imine-hydrobromide salt and this would be reacted with a methyl grignard reagent to form amphetamine.

The silver bromide from the Hunsdiecker reaction could be filtered out from the reaction mixture, the only "problem" is that nasty triethylamine hydrobromide. I have no idea how would that react with a grignard.
And also, grignard wouldn't be problem, because grignard reagents could be also prepared in trialkyl amines/not just in ethers.

Preparation of N-Formyl-DI,-phenylalanine

The method was a simplification of that of Fischer and Warburg. DL-Phenylalanine (82.5 g.) was dissolved in 100 cc. of 90% formic acid and boiled gently under reflux for two hours. An additional 100 cc. of formic acid was added and the solution allowed to cool. Formylphenylalanine then crystallized in readily filtrable form and was essentially pure (m.p. 165-166') after atering, washing with a little formic acid and drying in a desiccator over solid sodium hydroxide.

A second pure crop was obtained similarly after concentrating the filtrate by distillation to 75 cc. The yield was 83 g. (87%). Crops obtained by further concentration were slightly colored and contained some free phenylalanine or its formate as shown by failure to dissolve completely in hot
acetone. This material was best combined with similar material from other runs and crystallized from fresh formic acid. Working in this manner, the yield is nearly quantitative.

Crystallization from acetone left the melting point unchanged at 165-166' (uncor.)28 in agreement with reported values

Overby & Ingersoll "The Resolution of Amino Acids. I. Phenylanine and Valine" J. Am. Chem. Soc., 1951, 73 (7), pp 3363–3366 (attached)

L-N-Methylphenylalaninol (4a)

A 500-mL threeneck round-bottom flask equipped with a magnetic stirbar, reflux condenser, thermometer, and addition funnel was flushed with argon and charged with 250 mL of THF, 19.32 g (100 mmol) of N-formyl-L-phenylalanine(3 a), and 9.10 g (240 "01) of NaBH4 whereupon a vigorous gas evolution was observed. Then, a solution of 25.40 g (100 "01) of I2 in 100 mL of THF was added slowly and dropwise at a temperature of 25-40 OC. After the addition was complete, the flask was heated to reflux overnight. Excess reducing agent was cautiously destroyed by dropwise addition of 30 mL of MeOH at room temperature. The solvents were then removed in vacuo, and the residue was taken up in 100 mL of 20% aqueous KOH and the product extracted three times with 150 mL of MTBE.

After drying (Na2SO4), the extract was evaporated to a pale-yellow oil which was crystallized by the addition of 200 mL of hot n-hexane. Filtration and drying gave 84% colorless crystals which were recrystallized from 20 mL of toluene. Evaporation of both mother liquors and recrystallization of the residue from 5 mL of ethyl acetate yielded a second crop of crystals which was added to the first one to give a total yield of 73 % of 4a as colorless crystals: mp 71-74C (lit.Ig 68 OC); [a]D = +21.8' (1, EtOH) (lit.la +17.1° (2, CHCb)).

Marc J. McKennon, A. I. Meyers, Karlheinz Drauz & Michael Schwarm "A Convenient Reduction of Amino Acids and their Derivatives" J. Org. Chem., 1993, 58 (13), pp 3568–3571 (attached)

Preparation of the required chiral amine 14 (Scheme II) began with BOC protection of L-DOPA (9), followed by permethylation to give the ester 11. Lithium borohydride reduction yielded the primary alcohol 12, which was treated with methanesulfonyl chloride to provide the
rather unstable mesylate 13. The mesylate converted to urethane 15 upon standing; however, immediate removal of the BOC group followed by hydrogenation of the resulting trifluoroacetate salt gave pure 3,4-dimethoxyamphetamine (14). The proton NMR spectrum of the Mosher amide13 of 14 revealed only a single isomer.

Jonathan D. Bloom, Minu D. Dutia, Bernard D. Johnson, Allan Wissner, Michael G. Burns, Elwood E. Largis, Jo Alene Dolan & Thomas H. Claus, "Disodium (R,R)-5-[2-[[2-(3-chlorophenyl)-2-hydroxyethyl]amino]propyl]-1,3-benzodioxole-2,2-dicarboxylate (CL 316,243). A potent .beta.-adrenergic agonist virtually specific for .beta.3 receptors. A promising antidiabetic and antiobesity agent" J. Med. Chem., 1992, 35 (16), pp 3081–3084 (attached)