Reductive Amination of MDP2P with Al/Hg + Nitromethane

by Methyl Man


Since the appearance of Ritter's writeup of this method in Total Synthesis II, much discussion has taken place about it but, it has sometimes seemed, little has been clarified. This is due in large measure to the sensitivity of this reaction to even the most minor changes in its many variables. With this writeup, I hope to provide a clearer view of the method and to allow others to benefit from the hard-won experience of someone (not me) whom we'll call Mr. A. Ritter's original writeup, while inspiring, lacked details about the many nuances that, once understood, allow the amateur chemist to really understand this reaction's dynamics. Thus I have tried with this writeup to help the neophyte who has only physical observations and scant written material to guide him (although I suspect and hope that it may even help a few more seasoned cooks as well).

A photo-essay by Dr. Gonzo using this method can be found here.

The first thing I'd like you to look at is the array of interrelating variables in this reaction that make it so delicate. They are as follows:

  1. the thickness/type of the aluminum
  2. the consistency (i.e. flat, ground, etc.)
  3. the amount of HgCl2 used in relation to the amount of aluminum
  4. the addition rate of the MeNO2/MDP-2-P
  5. the size of the reaction vessel in relation to the scale of the reaction
  6. the ability to effectively stir the reaction
  7. the coldness of the water through the reflux condenser (yes, even that!)

The above factors are sort of submitted in an order of importance (#1 being most important), but in reality they are all inextricably related. I observed firsthand the trials and tribulations of Mr. A as he struggled to match up the correct combination of ratios and conditions that would allow a smooth, consistent reaction and predictable results every time. Finally, after lots of frustration, confusion, losses, and - in the end - a revelation, the perfect set of elements was hit upon and recorded.

The scale Mr. A chooses to perform this reaction on is half-scale to the scale in the Ritter writeup, which was 55g aluminum and 50g MDP-2-P. Therefore this writeup will illustrate the reaction on a scale of 27.5g aluminum and 25g MDP-2-P. The subject found for his own personal reasons that this smaller scale was much easier to manage (not the least of which is that even with a huge 4-liter separatory funnel, at this smaller scale it gets pretty filled up!). There's no doubt that the original larger scale can be successfully applied, although it would require adjustments in the glassware capacity, stirring method, and probably other elements.

Materials and Apparatus

  • 27.5g Reynolds Wrap Heavy Duty aluminum foil
  • 25g MDP-2-P
  • 20 mL MeNO2 of 99+% purity
  • 750 mL MeOH + 50 mL more for addition funnel + additional small amounts that will be needed later to thin the mixture
  • 400mg HgCl2
  • 2-liter 2-neck flat bottom flask
  • reflux condenser (400mm preferable)
  • 250mL or 500mL addition or separatory funnel
  • cooling setup (bucket, water pump, tubing, 1 bag ice)


  1. Weigh 27.5g of Reynolds Wrap Heavy Duty aluminum foil (NOTE: it HAS to be Reynolds and it must be the heavy duty stuff) and then tear it by hand or cut it with scissors into small rectangles approximately 1" by .75". Settle down with this task with a good CD or TV show because it is tedious and will take about 30-40 minutes.

  2. With a coffee grinder, "grind" these pieces of foil for durations of about 10 seconds. Fill the coffee grinder only loosely (about two thirds full - don't stuff it! That will adversely change the consistency of the ground foil). It will probably take about 4 to 5 "loads" in your grinder to do the whole 25g of foil, depending on the size of your grinder. (In actuality, the foil does not get "ground," but rather, each individual piece just gets compacted and compressed. If it is compressed too heavily, the inner surfaces of the foil nuggets may be rendered inaccessible to the Hg/MeOH solution, changing the timing of the amalgamation and maybe even causing an incomplete or failed reaction.) When properly done, the foil should be in gnarled little nuggets about the size of sunflower seeds (shell included) and should NOT look super-tight and small.

  3. Place a 3" stirbar in your 2L flat bottom flask and onto your stirplate. Add the foil nuggets to the flask and then proceed to set up your glass, support and clamps so that the reflux condenser and addition/sep funnel are securely affixed and your flask is well-centered on the stirplate (this will be critical when you begin to attempt stirring!). Also, prepare your cooling, i.e. attach the inflow and outflow tubes to the reflux condenser.

  4. Carefully add the 400mg HgCl2 to 750mL MeOH to a tightly sealable bottle and shake to dissolve all HgCl2. Set this solution aside.

  5. Combine the 25g MDP-2-P, 20mL MeNO2, and 50mL MeOH and pour them into the addition/sep funnel. Rinse your beaker (or whatever you used) with a tiny bit of additional MeOH to get the residual ketone and add it to this MDP-2-P/MeNO2/MeOH solution.

  6. Very slowly and carefully (w/gloves, glasses, long sleeves and a Hail Mary if you're Catholic), using a large funnel, pour the HgCl2/MeOH solution from step 4 down the condenser.

  7. Turn the stirring on full blast for a 5-second burst to intimately mix the solution and the foil. If you have prepared the foil as described above, it will easily stir. Give it a few more 5-second stirs over the next few minutes. I believe that doing this really helps facilitate the amalgamation process that is about to occur.

  8. After about 5 minutes or so, you will begin to see bubbles popping up on the surface of the MeOH solution. At first they will be tiny, like champagne bubbles. Then after a few minutes you will see them joined by larger bubbles closer to the size of those seen in boiling water. It is around this same time that the appearance of the aluminum will change from its normal shiny silver color and start to take on a dull gray look, accompanied by a gray cloudy look that begins forming in the MeOH. This is the magic moment when you want to begin dripping in your MDP-2-P/MeNO2/MeOH mixture. Set a drip rate of approximately 1 drop per second at this point and no faster. You can speed it up a bit later to accelerate the reaction if desired.

  9. Place about 3 lbs ice into your bucket. When you can feel exothermic warmth begin by feeling the outside of the flask, quickly add about 2.5 liters water to the bucket (or an appropriate amount to make very ice-heavy ice water) and plug in the pump.

  10. While monitoring the growing intensity of the bubbling amalgamation, turn on/off the stirring intermittently as you did earlier. This time it is to assure distribution of the added ketone/nitromethane in the reaction flask but also because the amalgamation seems to gain its vital momentum more effectively if given some significant blocks of time (meaning about 30 sec at a time) in between "stirring bursts." When the reaction is clearly starting to get vigorous and hot, crank the stirring to 10 and leave it on.

    NOTE: This is where you can take advantage of Mr. A's trial and error regarding this reaction's parameters. If you used the kind of foil specified, prepared it as specified, used no more than the specified 400mg HgCl2, and used a 2-liter and NOT a smaller flask, you can breathe easy knowing that the reaction is going to hum along nicely but will not get out of control, and will result in perfectly processed aluminum amalgam sludge. You may think that a 2-liter flask is oversized for this reaction, but that is precisely the point. The extra headroom in the glass provides a nice zone of "breathing room" for the reaction and facilitates good refluxing. I've seen this reaction get out of hand in a 1000mL flask, and it isn't pretty, believe me. Use the 2-liter.

  11. As the reaction progresses only a few minutes after the addition was started, you will observe that the aluminum is breaking up fairly rapidly. This is good, as long as you have the ketone/nitro mixture dripping in at a good rate of about 1drop per second. But be careful with the addition rate at this point, as a rate that is much faster than this could easily send the reaction into overdrive (not good). Your reflux should be unnervingly vigorous as the amalgamation really starts to pick up speed, with the MeOH literally pouring down out of the condenser. I know it's hard to believe, but this is what you want, this is good. I'm telling you, LOTS of trial and error came before this writeup. Trust me. You will also see sludge already starting to settle at the bottom and forming a ring on the glass around the top surface of the spinning mess. The consistency will get thicker by the minute. Add more ice to your bucket as needed.

  12. At this point you can sort of control the reaction rate by slowing down or speeding up the addition rate a bit. Of course the reaction is already barreling along, so you won't want to speed it up much. The concept here is that you want the addition of the ketone/nitromethane to be paced neck-and-neck, as it were, with the breakdown of the foil as it amalgamates and gets turned into sludge. In other words, you have to watch those two things and sort of adjust the addition so that they proceed at approximately the same rate. It's tricky, and imprecise, but with a little experience and intuition you'll get the hang of it. Sure, you could be lazy and just leave the addition at a steady 1 drop/second the whole time, but if the amalgamation peters out way before your addition is finished, and you find yourself adding your beautiful ketone to impotent sludge, don't cry to me. The addition should take about 40-45 minutes in total, and as it's finishing, the state of the aluminum should be about 95% broken down. In fact the reaction should by now (~45 minutes after addition was started) look like a really thick, steely-gray chowder with only minor small slivers of undissolved aluminum visible if any at all. You will probably even need to add an extra 20-30mL of MeOH down the condenser at this point (or before) to help it keep stirring effectively. This is no problem.

    A note about color at this point is helpful too. Comparing successful reactions to failed ones, I have observed that there is a distinctive color to the mixture early on that indicates healthy amalgamation and foretells a successful run. At a point maybe 30 minutes or so post-addition, the reaction takes on a color that I would describe as being "light steely gray with blue overtones." It is a hard thing to describe shades of gray, but I will try. It is a light shade, akin to the color of common gray sweat pants, but I'd say with a very slight suggestion of a blue hue in there as well. This is in contrast to what I saw in failures resulting from using too thick of aluminum and not enough HgCl2, where a dark metallic gray with definite green overtones (from unreacted ketone) was noted.

    NOTE: Another point I would like to make about the timing of the addition against the breakdown of the aluminum is that Mr. A found that there was a definite "spike" curve to the amalgamation reaction which was easily observed by watching the reflux rate. That is to say, there is a peak that it builds up to and then comes down from. At this scale, and using the exact materials described herein, that buildup to peak and subsequent slowdown occurs over approximately 25 minutes or so -- very fast. So at only about 20-25 minutes after you first started feeling the amalgamation heating up, it will have slowed to a reflux of about 2 drops per second, after having been at a peak with a reflux rate so furious it is a stream, not drops. At one hour and 15 minutes after you first started the addition, the reflux will have slowed to a very calm 1 drop per 2 seconds or so. Finally, when...

    1. the reflux has slowed to almost no reflux at all
    2. if you stop the stirring you do not see any small bubbles anymore
    3. no "uneaten" aluminum is visible and the solution is a thick, uniform gray soup,

    ...the reaction has essentially finished. It will reach this state at about one hour 45 minutes to two hours after addition was started. Nevertheless, you will leave it stirring happily for a total of three hours after the addition was finished to assure that the reaction has run its full course and the conversions that you desire have had ample time to take place. You could add a bit of external heat at this point but it's probably not even needed (Mr. A does it just to be sure). One reason I bring this all up is that there has been lots of talk about how this reaction needs 8 hours or 24 hours or even 36 hours to run! But those time frames apply only in cases where much thicker aluminum is used, and/or in variations using methylamine and not nitromethane. Mr. A was never successful in using thicker aluminum, and doesn't want to be! Why would someone want to make a reaction take any more time than it needs? Beats me! I'm mystified! The approach illustrated in this writeup optimizes this reaction to finish in 3 hours 45 minutes from beginning to end, and it probably doesn't even need that much time.

  13. If you chose to apply external heat, turn it off at 30 minutes before the targeted finish time. Otherwise you will have to wait an extra 30 minutes for it to cool for the next steps.

  14. When finish time has arrived, dismantle your setup, set aside your reaction flask, and make 750 mL 35% NaOH solution (750 mL H2O + 262.5g NaOH) and let it cool to room temp or below (safety glasses!).

  15. Into a separatory funnel no smaller than 2000mL capacity, pour your nice gray reaction mixture, being very careful to keep the stirbar from falling into the sep funnel and breaking it (that would be ugly). If your mixture is really thick, you may need to add small amounts of MeOH to thin it to a pourable consistency. This is perfectly fine. Wash the final residue out of the reaction flask with a few mLs of MeOH and add it to the funnel also.

  16. Slowly pour the NaOH solution into the sep funnel (gloves and glasses! no excuses!). That's right, don't dump it in wholesale. Basifying should be a gentle process. If you bully those molecules they may decide they're being disrespected and choose not to cooperate. Adding the NaOH will cause the mixture to warm up a bit as the very last bits of the aluminum are dissolved, which is fine. Swirl it a couple times and give it about 10 minutes to cool down to something closer to ambient temperature. That yummy stinky methylamine smell tells you that the reaction was successful.

  17. When the mixture in the sep funnel has cooled down, extract it once with 400mL toluene followed by once with 100mL toluene. These are the critical moments for your yield now, so you be sure to shake long and hard (at least 3 min) during these extractions (I don't have to tell you to vent do I?!). The toluene/product layer will of course be on top since toluene floats on water. Also, be sure to give the separations ample time to happen (at least 15 min); it is easy to tell when it's okay to separate because the interface of small toluene bubbles finally resolves and you have a nice clean line between the layers. If you like, do as Mr. A does and finish off with a final small extraction of 50-60 mL toluene just to get the last of the stuff.

    NOTE: Your extractions will contain a tiny amount of the base/metal/garbage from the bottom layer; this is inevitable but easily worked around in this way: when you have collected your combined toluene/product extractions in a bottle, chill that bottle in the freezer for 30 minutes or so. When cold, the garbage gets a lot less mobile and it is easy to decant the toluene away from it. Just be vigilant while pouring the last 50 mL or so and avoid letting that glob of crap rejoin the toluene. Yeah, you will lose the very last 2 or 3 mL, but that's life. Alternatively, you could filter it through a paper towel, but you will still lose the same amount when the towel absorbs it. Just get over it and move on!

  18. If you haven't already, drain the garbage layer out of your sep funnel into a storage bottle or something, and wash the garbage residue out your sep funnel with water.

  19. Wash the toluene/product 4 times (or more) in your sep funnel with 400 or 500mL H2O and a final time with 500mL of a saturated NaCl solution to remove any traces of solvated HgCl2.

  20. Dry your toluene/product solution with 30g of your favorite drying agent (MgSO4 recommended) in an acetone-cleaned, heat-dried bottle for no less than 30 minutes (Mr. A is superstitious so he lets it sit for an hour). Shake it a few times during this period.

  21. Filter the solution and gas it with that good ol' HCl bubbler setup. Be smart and use just enough muriatic (31% HCl) to wet the salt but not enough to make any puddles, and put a wad of drying agent wrapped in tissue paper in line somehow between the reaction flask and the tube leading to your pipette end. Weep with joy as a bumper crop of white precipitate crashes out of solution.

Expected yield: approximately 20-21.5g raw odoriferous product that will purify via careful recrystallization to 17-18g of beautiful snow-white MDMA! Ain't life grand?

One thing that I want to mention in hindsight here since I know it's probably going to come up, is the issue of added water.

It's been emphatically stated several times that some water is needed in this reaction for imine formation etc. etc. That may be true, but I know for a fact that it doesn't need to be added. That is to say, maybe there is some trace water in the MeOH Mr. A uses (Heet) that is fulfilling this function, and maybe it's even in all MeOH except expensive absolute anhydrous stuff. I don't know. But he doesn't add it anymore and it definitely doesn't hurt anything.

Before, Mr. A was adding about 25 mL, and maybe it was not a good thing. But he likes how it's going now.

Reference: Ritter, in Strike, Total Synthesis II, page 104-107

Reductive amination of MDP2P with Al/Hg using nitromethane as in situ amine source,
by Entropy (02-05-99)

While Charlie Brown and Linus were away at school, Snoopy decided to have some fun. He donned his Red Baron attire and jumped into his plane. He picked up the Big Bad Wolf from the forest who had just made MD-P2P (see MD-P2P via modified KRV thread) and the two of them headed for their laboratory along with the vial of MD-P2P.

Once in the lab, they were puzzled as to which way to go to produce their honey. Snoopy wanted to try the nitromethane Al/Hg procedure as outlined in TS II. The Wolf agreed, but insisted that they modify the procedure ever so slightly. They finally agreed and implemented the following procedure.

In a 2L: beaker, dissolve 1.5 grams HgCl2 in 1500 mL of distilled water. Then add 55 gm of aluminum squares (appx 1" sq) that have been cut from pie tins. Using a wooden chopstick (they had chinese for lunch) stir this mess around and let the Al amalgamate for appx 30 minutes. You should see effervensence and the less should turn a dirty gray color.

After the 30 minutes, dump the contents of the beaker into a strainer and rinse thoroughly with distilled water to remove as much of the mercury salts as possible. Shake the excess moisture off the Al amalgum and transfer to a clean 2L beaker, then add 850 mL of MeOH to the amalgum and set aside for a moment.

In another beaker combine 38 mL MD-P2P (you can use up to 50 mL with this recipe; they only had 38 mL to work with), 85 mL of 60% nitromethane racing fuel (60% nitromethane/40% MeOH), and 190 mL of methanol. [This is the equivalent of the amounts listed in TSII]

Now add the nitro/MeOH/MD-P2P mixture to the amalgum methanol solution. This does not have to be added slowly. They added it over a 5 minute period of time. They tossed in a 3" stirbar and placed it on a magnetic stirrer. The reaction started immediately. Within 7 minutes, the reaction was beginning to boil vigorously. Snoopy freaked out, ran and got an ice bucket with ice and placed the beaker on ice for a few minutes until the reaction was under control (a matter of 1-2 minutes) then the beaker was placed back on the stirrer and the reaction allowed to continue for 3 hours.

While the reaction was underway, the Wolf prepared 1 liter of a 35% NaOH solution and put it into the fridge to chill. Once the reaction time had elapsed, the sludge from the reaction was dumped into a large (2L) sep funnel along with the the liter of 35% NaOH solution . This was mixed well and allowed to sit undisurbed for about an hour.

NOTE: If there is unreacted Al amalgum, make sure the pieces of Al are removed or decant the sludge from the remaining amalgum. If there are unreacted pieces of Al amalgam left when the NaOH is added, this mess starts reacting again, this time as a methylamine generator, precisely the compound you were trying to avoid making (except for the in situ generation, of course) and there is not much you can do to stop it except hold your nose and wait or made a mad dash to the toilet.

After this time two layers were observed. The garbage bottom layer was removed and the top layer was transferred to a flask and the methanol evaporated leaving the impure amine.

The impure amine was then dissolved in 250 mL of DCM and washed with several portions of distilled water and finally with a saturated NaCl solution. To the remaining DCM solution was added 50 mg of MgSO4 and this allowed to sit for 30 minutes. The DCM solution was then decanted and bubbled with HCL gas yielding 23 gm (60%) of beautiful MDMA crystals. "Not too shabby for a first run" the Wolf muttered. Snoopy nodded with agreement.