MDMA via Tosyl Chloride Intermediate?


Has anyone ever heard of MDMA being produced by a Markonikov hydration of safrole and conversion of the alcohol to an alkyl tosylsulfonate (via rx with tosyl chloride) followed by Sn2 amination with methylamine? This synthetic pathway would be very similar to the bromination/debromination pathway that is recently in vogue, and might be even simpler. Plus tosyl chloride is dirt cheap ($23 for 1kg).


I try this procedure alot and it was much better than normal alkyl-halide process because the tosylate (and brosylate or nosylate ) don't give a lot of elimnation contrary to the alkyl-halide. I got the alkohol intermediate with oxymercuration of the allylbenzene and made the tosyaltion with tosylchloride/pyridine, then the SN2 in MeOH with a little THF ofr solubility purpose and RT . The yield and reaction time are:

  • >95%,20min, RT for oxymercuration
  • >95%,45min,RT for tosylation
  • 50-60%,5d,RT for SN2 with MeNH2-MeOH

The RT and polar solvent are very important because increase the temperatur favorise the elimination, decrease the polarity too. Anyway, this family of SN2 is favorised with polar protic solvent as MeOH (see the March). As I wrote under another topic, the tosylate are very good but there is even better leaving group they don't give any elimination and have better kinetic: the Triflate, but it's an expensive reagent.


For the tosylation of alkohol, the base is pyridine, because

  1. It neutralizes the HCl
  2. Very important: it form a complex with tosyl chloride which make the attack by the alkohol on the sulfur more easily .
For the hydratation of the alkene, acid medium is not good because the 1-aryl-2 propanol first formed is rapidly deshydrated to the stabilized isosafrole wich is hydrated to the 1-aryl-1-propanol. The result is:
  1. very poor yield in 1-aryl-2-propanol
  2. mixture of 1-aryl-2-propanol and 1-aryl-2- propanol which are not easy to separate ( must use distillation ) .

The oxymercuration process give only the 1-aryl -2-propanol intermediate without rearrangement in about 20min with >95% yield but HgCl2 can not be used, sorry. HgAc2 or Hg(NO3)2 or Hg(ClO4)2 or Hg(CF3COO)2 can be used. You must use Hg2+ (mercuric) and not Hg+ ion (mercurous). You can make HgAc2 from HgCl2+CH3COOH but you must purify it. Anyway HgAc2 can easily be purchased.


Thank you for the informative post. What originally got me interested in this method was the fortuitous discovery of 'triflate' anhydride in reasonable quantity. Have you used this leaving group in this reaction? Would you mind posting the details of your procedure? Ultimately, I suspect that the cost of this reagent makes it impractical for any scale up but for a small scale high-yield experiment it may be interesting.


I believe the only better leaving group than tosly is methanesulphonate from mesyl chloride, of course. The methanesulphonly group will probably provide yields 10% or so better in Sn2 substitution w/ anhydrous alcoholic methylamine or ethylamine soln. Methanesulphonyl chloride is a little more expensive than tosyl chloride however it is a liquid and therefor much simpler to handle than stinky irritating tosyl chloride.

On a side note safrol-2 mesylate reacted in 80% yield with excess benzylamine to make n-benzyl MDA which was easily hydrogenated at 30 lbs w/ five percent loading of 10%Pd/C catalyst to produce a total yield of 73% MDA from starting alkene. not too shabby! The 2 proponal was generated from methylenedioxyphenylacetaldehyde w/ MeMgI grignard.


Ritter : the oxymercuration process is very simple and easy to carry : RT , >95% yield , 25min reaction time ... The mesylate group is about the same than tosylate or nosylate or brosylate but the best known leaving group are triflate and nonaflate , tresylate is not so good it's about 400 time less reactiv than triflate but it is still about 100 time more reactiv than tosylate and analogs ... Conclusion : triflate is 4000 time more reactiv than tosylate and analogs ( mesylate , brosylate , nosylate ) . For a good explanation of the leaving goup see the chemist bible : "Advanced organic chemistry - Jerry March " chapter : nucleophilic substitution.


The tosylation must be conducted in pure pyridine . 11 mmole tosylchlorid is added slowly ( t<30°C ) to a stirred solution of 10 mmole alkohol in 10 ml pyridine . When tosylcl is added, the mixture is stirred for 30-40 min RT . Then the mixture is poured in 100ml 2N HCl, then tosyl is purified . Yield over 95% .


Just dug up some references for advancement of tosylate/mesylate esters as feasible, HIGH YIELDING synthetic intermediates to our beloved honey in an aqueous environment. The following is quoted from: Journal of Organic Chemistry 1988, 53, 4081-4

(R)- Tomoxetine Hydrochloride: A solution of phenyl-3-(2-methylphenoxy)-propyl methane sulfonate [the mesylate group is on the gamma carbon] (450mg, 1.45mmol) and methylamine (10ml, 40% in water) in THF (10ml) was heated to 65'C for 3h. After cooling, the solution was diluted with ether, washed w/ aq. sat. NaHCO3 soln. and brine, and dried with anhydrous K2CO3. After concentration a pale yellow oil was obtained which was dissolved in ether, bubbled w/ dry HCl gas [and you guys certainly know the rest] to produce a white ppt which was recrystallized w/ acetonitrile to yield title compound (400mg, 94%)

Thats amazing !!!! NINETY FOUR freakin percent from an aqueous MeAmine soln!!! Try achieving that with a halogen leaving group on our favorite alkene by cooking the stinky shit up in a pipe bomb w/ alcoholic MeAmine. You can't, if you are lucky 50% will be . Halogens blow as leaving groups compared to sulfonate derivatives as proved by this paper. It is such an advantage to be able to use good 'ol fashioned aqueous MeAmine compared to homebrew anhydrous alcoholic amine solns! The only drawback noticed immediately is the large excess of MeAmine employed by the authors. This shouldn't present a huge problem because the excess amine cooked off during the heating process can be collected by bubbling through HCl. The 65'C rctn temp is the bp of THF so excess amine will be liberated out of the top of the reflux condensor on rctn pot. A slow stream of N2 can be admitted through a bubbler in the second neck of the rctn. flask forcing the methylamine gas to be expelled out the top of the condenser into a beaker of HCl. Simply evap off HCl to obtain your excess amine back as MeAmine hydrochloride.

There is one other procedure for producing alkyl-methylamines from an alkyl mesylate using the exact same protocol listed above with product isolated in 96% yield! This proves the procedures high yields are reproducible, however both examples listed are performed on primary alkyl mesylates. Since we are working with a secondary alkyl mesylate yields may suffer a bit from steric hindrance during the nucleophilic substitution by MeAmine. Well actually, let me restate that.. my chem theory is getting a little weak.. In most if not all nucleophilic substition reactions in the literature compounds with a leaving group always have higher yields in nucleophilic substition rctns than a complementary compnd w/ a secondary leaving group. Therefore the mesylate in our case may not produce the 90+% yields quoted in the literature but it sure will be much higher than that produced by any halogen.

Siegfried: Excellent work in this area. Was wondering if you'd be kind enough to post the physical properties of the tosylate. Simple experience has proved that most tosylates are solids, however you're the only one who knows for sure. A melting pt. would be very useful. Any recrystallization solvent of choice?

Was an attempt ever made at esterifying the propanol produced w/ H2SO4? As a side note any tertiary amine can be used in a similar manner as pyridine to scavenge protons in the esterification rctn. Triethylamine was the amine of choice in the quoted article. Yields of 85% were recognized after several wasteful recrystallizations.

On the subject of hydration of alkene to alcohol, oxymercuration is obviously the most simple method considering rctn. time and yield. However soluble mercury salts just plain suck. It sure would be nice if the H2SO4 thing worked. Another possible synthesis may be a PTC rctn. between aq. NaOH and halosafrole. This is a well documented rctn. however conditions will probably have to be closely monitored to minimize isoalkene formation.

Finally, to sum everything up this is a major breakthrough because of the reactivity of amines to sulfonic esters in aqueous environment. Similar reactions in the past have usually reacted alkyl halides as leaving groups and fickle-to-make alcoholic amine solns with long rctn times or high temperature pipebomb pressure vessels. Not very desirable when compared w/ a 3hr STP reflux. Reported yields are also very poor w/ halide exchange rctns. Comment?