Notes on the synthesis of MDMA precursors from Eugenol

[ Back to the Chemistry Archive ]

Eugenol demethylation (by Drone 342)

I think I see where your cencern about the use of AlCl3 stems from, and I wanted to clarify this issue, if I haven't already (sort of preliminary notes on the forthcoming report.) In TSII, you describe how AlCl3 can destroy safrole, and this is to be expected. However, that same phenomena that will wreck safrole is precisely what is being taken advantage of in the case of demethylating eugenol.

Anydrous AlCl3 is a relatively strong Lewis acid. It bonds with the lone pairs of electrons on the oxygens in the ethers. When this complex is hydrolyzed, the ether is cleaved. When you do this to safrole, you can definately expect a mess, since the methylenedioxy ring gets cleaved, and things can get even more out of hand from there. In the case of eugenol, this cleavage is exactly what we're looking for.

The use of lewis acids to cleave aryl methoxy groups has a long history of success. Its by far the most gentle, most widely appilcable procedure for doing this. The following is a list of demethylations employing aluminum chloride or boron tribromide on a wide variety of different compounds. Its interesting to see that this reaction is so selective, that by carefully measuring out the amount of Lewis acid, selective ether cleavages can be undertaken successfully:

(demethoxylation of N-allyl norcodeine derivatives; can't get pickier than that)
J.Amer.Chem.Soc.; 75; 1953; 4963, 4966;

(cleavage of 3-methoxy allylbenzene)
Helv.Chim.Acta; GE; 61; 1978; 401-429;

(cleavage of allyl-containing heterocyclic opioid -related compound)
J.Med.Chem.; EN; 21; 1978; 423-427;

(another convoluted methoxylated polycyclic arene, with an allyl dangling randomly from another ring)
J.Med.Chem.; EN; 29; 4; 1986; 531-537;

Eugenol to MethylVanillylKetone (by Psychokitty)

This is really weird, Drone, because I was just about to post why it is that AlCl3 CAN'T be used to demethylate eugenol. But my reasoning was based on a section that I read in a very prestigious organic chemistry textbook wherein, as one example, tetralin was being synthesized using a butylbenzene with a chlorine atom dangling from the very end of the butyl aliphatic side chain. This reaction, according to the text, is applicable to any situation where the starting compound has a similar structure (one having a susceptible group at the end of the side chain, such another halogen, an alcohol group, or, would you believe it, a double bond!). So the AlCl3 catalyzes in Friedel-Crafts fashion the cyclization of the side chain, which in this case, forms the desired tetralin. With this new information, I was about to suggest that using AlCl3 to cleave eugenol's methoxyl group would only actually serve to cyclize the allyl side chain to form a substituted indan ring. This would be good if one could make use of such a thing, but for the purposes of demethylating eugenol... no, I don't think so. But as I'm very happy to see, I was wrong. After all, what's with the 3-methoxyallylbenzene cleavage? If AlCl3 ether cleavage can work effectively on that molecule, I see no reason why it can't work on eugenol. Perhaps the reason AlCl3 does work in this case is because of the 3 carbon allyl side chain in eugenol in contrast to the 4 carbon butyl side chain in the above tetralin synthesis example. Maybe the double-bond is not the best candidate for strong effecient reactiveness in the aforementioned cyclization reaction. Who knows and who cares, because all that matters is that the damn thing works!

I've come up with a protocol that obviates the typical eugenol to allylpyrocatechol route. I figure that the best way to go about tackling this problem is to try to use a simple, readily available, easily synthesized cleaving agent that is generally safe to use, is predictable, that works in an efficient fast manner producing high yields of product while having little or no need of various other chemicals or solvents, that provides an easy work-up, and that can be applied in a manner that has already been detailed in the literature. My method meets all of the above criteria. As an added plus, it bypasses the use of any controlled precursors (unless you want to invoke the analogues law, which won't apply, I'll later explain). It makes use of all the good work that has been poured into the hive by dedicated bees, and in turn, offers an easy-to- follow scheme that is all-too likely to work. Pay close attention all you busy bees who already have basic experience with the standard safrole to MD-P2P to MDA route. This method is so straight-forward, that you can begin experimenting with it as soon as possible, because unlike all other ether cleavage reactions, this one is pretty hard to fuck up, unless, like Strike says, you're a total fucking idiot. Excited? I am. Now for the goods:

I decided that the first step of oxidizing eugenol to methylvanniyllketone (MVK) should be an already established one. I have no doubt that using the Wacker to effect this change would work better than well. It's just that to give my method more credibility, I need to have a proceedure that has already been proven to work.

First, eugenol needs to be isomerized to isoeugenol. I've read that the standard EtOH/KOH method is not well-suited to eugenol in that the reaction produces low yields of isoeugenol. If one were to just use the Wacker on eugenol, then this step could be skipped, but for the sake of accurracy, I'm going to have to include it.

This method was originally applied to synthesizing isosafrole from safrole. It is a super method that I have no doubt will work equally well on eugenol to make isoeugenol.

Eugenol (500g), iron pentacarbonyl (2.5g), sodium hydroxide (1.6g) were mixed in a 1Lt flask equipped with stirrer, thermometer and condenser. The well stirred reaction mixture was heated to 110degC at which temperature a vigorous reaction commenced, causing the temperature to rise to 1890degC in 6 minutes. After cooling the mixture, 250 ml of 2N acetic acid were added. The organic layer was separated from the aqueous phase and washed with brine to neutrality. After drying and evaporation of the solvent the mixture was distilled from a Claisen flask and gave 485g (~97%) of isoeugenol.

The next step uses Strike's sentimental favorite. There are two variations presented, though: a.) the peroxidation of isoeugenol to form the intermedate diol which is dehydrated to MVK; and b.) the reaction of peracetic acid with isoeugenol to form the epoxide, which is then hydrolyzed to the diol, and lastly dehydrated to MVK.

Taken from UK Patent 2,059,955 A.

MVK from isoeugenol (through process)

A solution of 30% aqueous hydrogen peroxide (9ml,85.5mm) and formic acid (16ml,88%)is added to a solution of isoeugenol (8.1g,50mm) in formic acid (4ml). The reaction mixture is stirred at 35-40degC under nitrogen atmosphere for 3 hours. The resulting 1-(4-hydroxy-3-methoxyphenyl)-propane-1,2-diol- monoformate is treated with 10% aqueous sulfuric acid (125ml) and toluene (125ml). After refluxing with mixing for 6 hours, the reaction mixture is cooled to room temperature, and the toluene layer separated. The aqueous layer is extracted with fresh toluene and the toluene layers are combined, washed with saturated aqueous sodium sulfate, dried over anhydrous sodium sulfate and concentrated to give MVK in 48% yield. Purification can be done one of two ways. The MVK can either be distilled under reduced pressure or it can be separated by stirring with 5%NaOH in which it forms the sodium phenolate which is soluble in the aqueous phase. The solution is then extracted with solvent to remove the non-phenolic compounds and then acidified to release the MVK.

This next method uses the through process also, and makes use of a buffer much like the process described in CA (1975)82,72640.

MVK from acetyl isoeugenol

A mixture of isoeugenol (8.21g,50mm), acetic anhydride (5.62g,55m) (feel free to substitute this very naughty chemical with the less proscribed triflouroacetic anhydride), and anhydrous sodium acetate (0.41g,5mm) is heated at 100-105degC under nitrogen atmosphere for 2.5 hours. The reaction mixture is cooled and diluted with 65ml of toluene before a solution (11ml) of 38.6% peracetic acid and 0.85g of sodium acetate in acetic acid is added slowly. After heating at 50-60degC for 2-3 hours, the reaction mixture is cooled to 20degC and treated with aqueous sodium bisulfite (2.8g) to destroy excess peracetic acid. The entire mixture is mixed with 8ml of toluene, 50ml of 14% aqueous sulfuric acid, and is heated with stirring to reflux for 12.5 hours. After cooling to ambient temperature, the toluene layer is separated, and the aqueous layer extracted 3 tms 20ml toluene. The toluene layers are combined, washed with saturated aqueous sodium sulfate, dried over anhydrous magnesium sulfate and concentrated in vacuo to give MVK in 85.3% yield.

The next step is standard. There are many good various ways to get to the amine. Personally, from the established list of methods, I'd pick the reduction using sodium cyanoborohydride. But for the sake of novelty, I'm going to include another more promising method that has yet to get much attention, but which will eventually supplant sodium cyanoborohydride as the #1 reagent used to get to MDA.

This next method is very clean and high yielding, and uses two cheap reagents (sodium borohydride and titanium(IV)isopropoxide) which are unwatched, readily available, and safe.

A mixture of MVK(10mm), titanium(IV)isopropoxide (5.7g,20mm), ammonium chloride or methylamine (20mm), and triethylamine (2.5g,25mm) in absolute ethanol (10ml) was stirred at room temperature for 20hr. Sodium Borohydride (.95g,25mm)was then added and stirring was continued for a further period of 20hrs. The reaction mixture was quenched with 30ml,2N)of ammonium hydroxide, the resulting inorganic precipitate was filtered and the aqueous solution was extracted with dichloromethane (50ml tms 3) to separate the neutral materials. The aqueous solution was then made alkaline by addition of 10%NaOH (pH=10) and extracted with dichloromethane (50ml tms 3). The combined dichloromethane extracts were dried (K2CO3) and concentrated in vacuo to give 2.2g (75%) of 3-methoxy-4-hydroxyamphetamine (MHA?) or 3-methoxy-4-hydroxymethamphetamine (MHMA?).

Information for the above posts was taken from the following references:

JOC 1995, 60, 4928-4929 "Reductive Alkylations of Dimethylamine Using Titanium(IV)Isopropoxide and Sodium Borohydride: An Efficient, Safe, and Convenient Method for the Synthesis of N,N-Dimethylated Tertiary Amines"

SYNLETT, Oct. 1995, pp1089-1080 "An Efficient, Safe and Convenient One-Step Synthesis of B-Phenethylamines via Reductive Amination Reactions Utilizing Ti(OiPr)4 and NaBH4"

The next step is the most ground breaking of all. Strike, you're always complaining about how HBr has always proven to be less than ideal when it comes to ether cleavages forming low yields of product with the unfortunate concommitant formation of loads and loads of tar. So naturally, you have little expectations of HBr's efficiency and plenty of doubts that it will ever be appied to any worthwhile degree to the formation of that oh-so-desirable catechol species we're all after. Well, I think I've found exactly what you've been looking for: A method whereby ether cleavage is effected rather cleanly, in high yield, simply, using a readily available reagent, in a short amount of time. Here we go:

One part of 3,4-MHA or 3,4-MHMA is refluxed for 1-2 hours in 5 part of 48% HBr to yield 80-90% 3,4-dihydroxyamphetamine (DHA) or 3,4-dihydroxymethamphetamine (DHMA). And that's it! The solution is completely homogenized and is low enough in volume to make scaling up cheap and easy. And the HBr can even be recycled too.

Next part: The HBr solution is distilled at room temperature or under reduced pressure. Care is taken that the heating bath never exceeds 150degC as the melting point of the amphetamine salt will be somewhere above 150degC area (I don't have the exact melting point yet). After all the acid has evaporated, there is left a colored residue which is dissolved in alcohol, decolorized with Norit (activated carbon), and precipitated by adding ether. The salt should be sufficienty pure to use in the next step.

Information for the above post was taken from the following references:

Fred W. Hoover and Henry B. Hass. JOC, 1947, pp.501-505 - "SYNTHESIS OF PAREDRINE AND RELATED COMPOUNDS"


JACS 54, p271 (1932) & 60, p465 (1938)

One that I haven't check out but is applicable is JCS 1951, pp.2248-52. Someone get the article for me please so that I can know what the yield is of 3,4-DHA from 3,4-DMA hydrolyzed under reflux in HBr.

Now for the next step. The methylenation reaction. Admittedly, this is the one step where I feel all of us are wading in the dark. It's not that performing it would bee especially difficult. It's just that no one knows how effective, in terms of yields and EXACT reaction conditions, it will be. My interest was in choosing a methylenation reaction that would allow one to proceed smoothly from the previous reaction into the next and last step towards the final product. My needs have led me to the previously published hive proceedure using a PTC, excess dichloromethane, water, NaOH and in this case, the catechol-amine which is then reacted for cal. 5 hours at 70-80degC under 2.4 atm of pressure. Now before I move on, I'd like to comment on the concern that many bees seem to have about the effectiveness of using an "unprotected" amine in the methylenation reaction. In regard to this, I'm stumpted as to why anyone is concerned at all. Perhaps they just don't understand the reaction mechanism. Allow me to enlighten and elaborate. The methylenation reaction with which all of us are so familiar is formally called the Williamson Reaction. What is occuring at the molecular level when it is carried out is as follows:

NaOH, a base, reacts with the catechol (an acid) forming the sodium catecholate, a salt through a simple neutralization reaction. The halogens in the dichloromethane react with the sodium catecholate by swapping (substituting) their halogens with the Na atoms to form NaBr and the desired methylenedioxy compound. And that's it! With PTCs, the reaction is made to go faster and in higher yield. But the reaction mechanism remains the same. With dipolar aprotic solvents, the SN2 reaction (don't ask) occurs faster, but again, the reaction mechanism remains the same. In the methylenation reaction using KF, the KF handles the task that the NaOH normally would, except it does so better. Other than that, everything else stays the same. It is common knowledge for those who understand the nature of this reaction that because of the way it occurs, many functional groups are tolerated during the course of the reaction...

Continuation from above section - Psychokitty

Okay, sorry, sorry, sorry. The correct temperature in the isomerization step is 180degC. In the reductive amination step, use methylamine HCl, not methylamine base. And in regard to the explanation of the Williamson Reaction, it should be NaCl and NOT NaBr that is formed as a byproduct. Sorry about the little mistakes. If any of you happen to catch any more, I'll be happy to make the corrections . . . Now back to where I left off yesterday.

Ah, yes! Regarding the feasibility of using the free base amine in the methylenation reaction. Well, there's nothing more to write about beyond that which I posted yesterday. By every rule in the book that I know of, a catechol with an amino function incorporated into it should present no problems when used in the Williamson Reaction. If I'm wrong and any of you bees out there can explain why, then believe me, I'm all ears. But, again, I'm concerned aboout my method not being as all-inclusive as everyone should expect it to be, so for the sake of accuracy, I'm going to add one additional step before I proceed to the actual methylenation: The protection of the amino group.

First, isolation of the amine free base is in order. Catechol-amines, from what I can tell, are pretty tricky when it comes to using standard basifying techniques. Why? Well, in this case, if you just basify with NaOH as usual, you won't get the expected free-base, but instead, will be left with an homogenous base solution with the following dissolved in it: NaOH and the sodium catecholate of 3,4-DHA or 3,4-DHMA. So then how does one go about liberating the free base? Well, I've read time and again that phenols are insoluble in sodium bicarbonate solutions. Why? I don't really know, actually. But what I do know is that sodium phenolate is decomposed by atmospheric carbon dioxide, so it's my guess that the carbon dioxide that is essentially part of sodium bicarbonate is what contributes to the retardation of the formation of the sodium phenolate, making phenol insoluble in sodium bicarbonate solutions. Who knows? I sure don't. But I would appreciate any comments from my fellow bees regarding this perplexing issue. Anyway, back to the basification step: It seems to me that a sodium bicarbonate solution of about 25% might be sufficient to neutralize the amine-hydrobromide, thus liberating the free-base amine. I suppose this step could be used immediately after the ether group has been hydrolyzed, simply by cooling the HBr solution and then basifying right of the bat. However, I think it more prudent and conservative to evaporate the acid first, and then basify. In the end, though, its up to the experimentor.

Another alternative to basifying with NaOH would be to use concentrated ammonium hydroxide. In this way, there would be no Na ions in the solution to contribute to sodium catecholate formation. After basification, simply extract, dry, and evaporate to get the desired free-base. Weigh the amount thats there and then go on to the next step...

The information contained in the next post was taken from the following reference:

J. Med. Chem. 1993, 36, 3700-3706.
"Synthesis and Pharmacological Examination of Benzofuran, Indan, and Tetralin Analogues of 3,4-(Methylenedioxy)amphetamine"

Synthesis of N-(Triflouroacetyl)-1-(3,4-dihydroxyphenyl)-2-aminopropane. To an ice-cooled solution of 16mm of 1-(3,4-dihydroxy)-2-aminopropane (your isolated free base from the last step) dissolved in 100ml of dry CH2Cl2 was added 2.0g (19.7mm) of triethylamine. The mixture was stirred for 10 min and 9.8 (49mm) of triflouroacetic anhydride in 50ml of CH2Cl2 was introduced dropwise to the reaction vessel over a 10-min period. The reaction was allowed to warm to room temperature. After 1.5h the solvent was removed by rotary evaporation and the oily, yellow residue was taken up in ether. The organic phase was washed with H2O (2 tms 25ml), 2N HCl (25ml), 5% NaHCO3 (25ml), and brine and then dried over MgSO4 and filtered through Celite. After complete removal of solvent in vacuo, a quantitative yield of white solid was obtained (may not be a solid in this case). This was recrystallized from ethyl acetate-petroleum ether to afford ~89% of the desired product as fluffy white crystals.

For those interested, the above reaction is not vigorous enough nor at a high enough temperature to create problems with attendant formation of the 3,4-triflouroacetyl ether through reaction of triflouroacetic anhydride and the catechol. Just thought you'd like to know.

Anyway, so now we have our "protected amine". On to the methylenation reaction.

Although the reference I got this from didn't indicate so, my impression is that there should be stirring of the mixture throughout the course of the reaction.

Using the unprotected amine:

100ml (1.56m) of methylene chloride, 0.02m of hexadecyltributylphosphonium bromide (PTC) and 200ml of water were placed in an autoclave, and a total of 0.2m of 3,4-DHA-HBr or 3,4-DHMA-HBr and 33 or so grams of NaOH flakes (Note 1) in 30ml of water were added in stages at a temperature of 70degC (Note 2). The pressure increased to a maximum of 2.4 atm, and the reaction was continued for 4h. After this time the reaction mixture was cooled to ambient temperature, the organic phase was separated and the excess methylene chloride was recovered by distillation. Yield is ~70%.

Note 1: The original amount of NaOH was 27.6g [0.6m] but because of the amine is in its HBr salt form, more NaOH will be needed in order to neutralize it.

Note 2: The reference I used indicated that the temp was 700degC. Naturally, I assumed it was a typo.

Information for the above post was taken from Rhodium's site under the guest.eugenol.txt. The original information came from the following British patent spec.: 1,518,064.

So now where do we go? Well, first, let's determine where we stand.

If the unprotected catechol-amine is used directly, then through the course of the reaction, the NaOH reacts with amine hydrobromide to first form the free base, and then the free base reacts with the excess NaOH to form the respective sodium catecholate. The sodium catecholate then reacts with the methylene chloride to form the desired MDA or MDMA product.

Unlike typical methylenation reactions, this one takes place through the heating and agitation of two immiscible phases. It is the presence of a PTC which optimizes this reaction by carring the normally insoluble ions from the polar phase down into the organic non-polar phase. This, of course, expedites the rate of the reaction. At the reaction's end, there will be two immiscible phases. The bottom is the methylene chloride layer having dissolved in it the PTC salt, possibly some dimer, and hopefully, the desired MDA or MDMA. The upper polar phase will contain NaOH and possibly some unreacted sodium catecholate. The layers are separated and the methylene choride layer should be washed several times. Crystallization can be commenced at this point, but it would be best to back-extract the amine from the methylene choride with either 15% HCl or 10% H2SO4, separate the organic layer, and then basify the acid solution to liberate the amine. Reextraction with a suitable solvent and subsequent crystallization may be the next way to go. Otherwise, reextraction, then solvent separation, drying, and then evaporation, followed by vaccum distillation of the product, is in order. Expected yield ~70%.

If the protected N-triflouroacetyl-catechol-amine is used, then do not increase the amount of NaOH in the reaction; instead, use the original 27.6g (0.6m). Proceed as instructed above. After the completion of the reaction, isolate the N-triflouroacetyl -MDA or -MDMA through evaporation of the methylene chloride and proceed to the following deprotection step:

A solution of 5.12mm of protected amine in 110ml of 2-propanol (IPA, isopropylalcohol, rubbing alcohol)was vigorously stirred while 10ml of 2N KOH was added, and the mixture was heated at relux for 5h. After cooling, the solvent was removed by rotary evaporation. The residue was taken up into 150ml of 3N NaOH and the aqueous solution was extracted with CH2Cl2 (4 tms 50ml). The organic fractions were combined and then extracted with 4 tms 50ml of 3N HCl. The acidic aqueous extracts were combined and then basified 5N NaOH to pH 11 (external damp pH paper) while cooling on an ice bath. The free amine was extracted into CH2Cl2 (4 tms 25ml)and the organic phase dried (MgSO4), filtered through Celite, and concentrated on the rotary evaporator. The residual yellow oil was dissolved in 15ml of anhydrous ether, and the hydrochloride salt was formed by the addition of 6ml of 1.0 N HCl in anhydrous ethanol. After removal of the volatiles by rotary evaporation, the resulting white solid was recrystallized from ethanol-hexane to yield ~78% of white crystalline MDA or MDMA.

Information for the above process was taken from the aforementioned J. Med. Chem., 1993, 36, 3700-3706.

A few final notes of worth:

  1. During the hydrolysis of the vanniyllisopropylamine in the refluxing HBr solution, there will be considerable MeBr evolved. Therefore, steps should be taken to efficiently vent this poisonous by product. Those with the knack to do so, can pump the MeBr gas into a cooled alcoholic solution for later reaction with hexamine and NaI to form methylamine-HCl.

  2. To the voters: During your deliberation on the merits of this method, remember to remain focused on the efficiency of the ether cleavage. Strike said he was mostly going to base his final decision about which method is best by only generally considering the catechol-forming step. So it seems like all other proceedures, such as those detailing the formation of the ketone and amine, are to be considered as nothing more than window dressing. However, they are included in my post to make the method complete. Remember, ether cleavage of eugenol and MVK followed by methylenation to form safrole and MD-P2P, are methods far from established and rife with guesstimations and uncertainties as to the length of reaction time and the likely but questionable yields. Those methods are, at best, far from established and would require numerous tests to determine exact and accurate parameters. My hydrolysis step, however, is quite established. And for all you gripers out there, I know what you're thinking: None of the examples that I gave actually use vanniyllisopropylamine in the hydrolysis step. Well, yes, that's true, but I did find relevant experimental information that allows one to make a very accurate educational guess as applicabilty of the HBr hydrolysis step. I mean, come on! No one is going to tell me that because vanniyllisopropylamine isn't 4- or 2-methoxyamphetamine or 3,4-dimethoxyamphetamine, that it isn't going to react in the same way! That instead, one is going to go from 80-90% yield, down to 20% with the unfortunate concommitant formation of loads and loads of tar! To suggest such a thing is ludicrous! Furthermore, if it were likely, then I would say that we're all in a lot of trouble. Why? Well, if you think MY example needs to be more substantiated, then consider how far off the mark the other TOTALLY untested methods (such as the hydrolysis of eugenol using god-knows-what historically unappied reagent) really are.

  3. This is for those of you who think that refluxing the amine salt to form the catechol-amine in concentrated HBr is nothing more than a recipe for the formation of a butload of tar. I ask you this: What is presently the most popular way of synthesizing d-meth from ephedrine? Answer: The red-P/HI reduction! And how does this reduction occur? By refluxing the ephedrine in concentrated HI! How long is this reaction? Fucking 24 hours! What are the yields and purity of product typical of this reaction? High! So what's the problem with refluxing the amine in concentrated HBr for just one hour? THERE IS NO PROBLEM!

  4. For those of you who would prefer to use the cheaper and more readily available concentrated HCl to effect the ether hydrolysis, refer to the following: JACS 54, 271 (1932) & JACS 60, 465 (1938)
    You'll have to carry out the reaction in an enclosed vessel under pressure. Yields and reaction time are about the same. Don't know how you're going to safely vent out all that MeCl byproduct, though.

  5. One other interesting thing. It seems pretty clear now that it will be unfeasible to sythesize ether substituted amphetamines through the HI reduction of their ephedrine counterparts simply because ether cleavage is going to occur much much faster than reduction of the beta-hydroxy group. Wierd, huh?

Continuation from above section - Psychokitty

Taken from JOC Vol.23, pp.1783-1784.

Looks like your idea about hydrolysing MVK to form the dihyroxyphenylacetone was right on the mark, Mr. Drone. The following is a demethylation sythesis using concentrated HI as the cleaving agent. Yields are a little low, but hey, you wisely suggested using a PTC to avoid that problem. The following sythesis uses 2,3-dimethoxypropiophenone as its starting material. It's not too different from MVK, so using MVK in its place should proceed smoothly.

2,3-dimethoxypropiophenone (2.8g) was refluxed with hydriodic acid (11g) and an equal volume of glacial acetic acid for 6 hr. The reaction mixture was then poured onto ice and left over-night. The precipitated product was filtered off, dissolved in benzene, and the dark solution treated with charcoal. To the filtrate after concentration, a few drops of petroleum ether (40-60degC) were added, whereby 2,3-dihydroxypropiophenone separated out. It was recrystallized from petroleum ether (40-60degC) in pale yellow crystals, mp 53degC, yield 41%. It gave a green color with alcoholic ferric chloride solution which changed to red on the addition of sodium carbonate solution.

As for the last part of the above scheme, I'm not sure that it would definitely apply. Depends, I guess, on whether or not 3,4-dihydroxyphenylacetone is a solid at room temperature.

Next on the hit list is another example where a lewis acid is employed - AlBr3 - as the cleaving agent. According to a limited review I read recently, its even better than AlCl3 for that purpose. Again, I don't know if there would any side reactions due to eugenol's double-bond. But to me, it doesn't matter because this method requires the use of way too much toxic solvent (acetonitrile or carbon disulfide) for converting such a paltry amount of eugenol.

Somewhere in the article, the authors rambled on about how they were very selective and careful about which demethylation they chose to employ. And, although the method is not all that great, I enthusiastically applaud these guys because in most of the other papers I've read, nothing more than basic information is ever offered. You all know what I mean.

The final product on this one is a benzaldehyde called Atranol. Must have been a much-needed compound back in 1948. Anyway, I guess this means that AlCl3 complexes with ketones but not with aldehydes. Who knows. Who cares. Here's the goods on the proceedure:

To a solution of 5g of atranol dimethyl ether in 250ml of carbond disulfide in a 500ml round bottomed flask fitted with a mechanical stirrer, 22g of aluminum bromide (3 moles per aldehyde) in 250 ml of carbon disulfide was added quickly with stirring. The addition complex precipitated as a red gum. After stirring for one hour the carbon disulfide was decanted into a separatory funnel and 100g of crushed ice, 150ml of 3N hydrochloric acid and 200ml of ether was added to the residual gum in the flask and stirred until it was completely dissolved (one or two hours). The carbon disulfide in the separatory funnel was washed with 3N hydrochloric acid, the carbon disulfide layer was discarded and the acid layer added to the mixture in the flask. The ether layer was removed and the aqueous layer extracted with two 200-ml portions of ether.

The combined ether solutions were extracted with three 50-ml portions of 1 N aqueous sodium hydroxide. The atranol was precipitated from the alkaline solution by addition of concentrated hydrochloric acid. The product was filtered, washed with a little water and recrystallized from 125 ml of boiling water (Norit). Yield (average of twelve runs), 2.9g (70%)of slightly yellow product of adequate purity for the next step.

Information for the above post was taken from the following citation: JACS Vol. 70, pp.2120-2122 (1948)

Now we come to the grand prize offering. I must say that I truly love the potential the next method has to offer. The reaction is quick, the yields are high, only one reagent is used to effect the cleavage which it seems can also be recycled, no toxic by-product seems to be formed, and best of all, 4-hydroxy-3-methoxy-propiophenone is used as the starting compound! Short of finding a published proceedure detailing the ACTUAL ether cleavage of EUGENOL, things can't get any better than this, folks!... Oh, what am I saying? Of course it can!

The proceedure detailed in the following post was taken from several different sources. For those that want mechanistic proof of the method, refer to Fieser and Fieser's "Reagents for Organic Sythesis, Vol 1". I don't remember what the page numbers were. Just look up "pyridine hydrochloride" in the index and turn to the relevant pages. Easy enough. Let's proceed.

Preparation of 4-propionyl catechol. Taken from JOC Vol 26, pp.2401-2402 (1961)

A mixture of 5g of 4-hydroxy-3-methoxypropiophenone and 10g of redistilled pyridine hydrochloride was gently refluxed (200-220degC) for 10 min; after cooling and addition of dilute hydrochloric acid, the precipitate formed was washed with water and recrystallized from aqueous ethanol, giving 4-proprionylcatechol in 80% yield.

Here's another way to proceed after the reflux step. Details to the process were taken from JACS Vol. 79, pp.147-148 (1957).

After cooling, the fused mass was crushed in 15ml of 5% hydrochloric acid and the mixture extracted with ether. The etherial solution was washed with water, extracted with N potassium hydroxide, the combined basic extracts washed with ether, then acidified with concentrated hydrochloric acid and stored over-night. There was obtained the final product as colorless crystalline material.

And there's even another way to go - Taken from JOC Vol. 27 pp.4660-4662 (1962).

After the flask has cooled, the reaction mixture was dissolved in ether and diluted with water. After three extractions with ether, the combined extracts were washed with dilute sodium chloride solution and dried over anhydrous sodium sulfate. The solution was evaporated almost to dryness and the residue crystallized from a benzene-petroleum ether mixture to give the purified desired product.

I'm sure the pyridine can be recycled but I don't really know exactly how right now. That's a proceedure I'll outline later.

Oh, I forgot to mention something! If it turns out that this reaction is incompatible with eugenol, then what needs to be done first is the oxidation of eugenol via the Wacker to form MVK. Then one can proceed to the above demethylation step. MVK, I'm quite sure, can be used in place of 4-hyroxy-3-methoxypropiophenone as there is only one slight difference between the two. Namely, where the carbonyl group is located on the propane side-chain.

Isomerization of Eugenol to Isoeugenol

by Psychokitty

500 g of clove oil (undistilled; about 98% eugenol) was put in a 1000 mL erlenmeyer flask along with a 3" stir bar. Added 150 g of KOH (calculations required 200 grams but only had 150 grams available) and 25 grams of PTC (hexa-decyl-tributylphosphonium bromide).

Started to heat the reaction using a hotplate/magnetic stirrer. Used a standard thermometor to measure the increase in temperature. Did not make use of a reflux condenser nor did I apply reduced pressure to effect this reaction.

Tried to stir mess while the temp was rising. Even with a stong magnetic stirrer, this was virtually impossible (the solution was full of KOH which after a while made the surrounding solution look opaque brownish-black).

Temp finally rose to about 100 deg C or so and then WHOOOOSH!!!! An exothermic reaction suddenly kicked in making the temp rise quickly to 125 degrees C. And before I knew what was going on, the KOH was being absorbed with the occurance of a simultaneous deposition of potassium eugenolate (brown color).

As the temp rose to about 135 deg C, the potassium eugenolate began to melt and by 140-145 or so degress C, it was completely melted making the potassium eugenolate mixture homogenous (black opaque solution). The smell of eugenol was very powerful (but not terribly unpleasant) so to keep it from evaporating easily, a damp paper towel was crumpled up and fitted loosely into the mouth of the flask. This application worked very well.

About 30 seconds later, the stirrer began working and when the temp reached about 150 deg C, the solution started to boil. YES, IT STARTED TO BOIL. Stirring and heat were maintained for fifteen minutes at which point a somewhat unpleasant (but tolerable) smell filled the air, which promted me to remove the flask from the stirrer/hotplate to cool (used a fan).

As the temp reduced to about 135 deg C, the solution began to crystallize and by 125 deg C or so, solid potassium eugenolate was everywhere in the flask (brownish tinge). It was so solid that I had trouble removing the thermometer.

The solution was heated up to 150 deg C again slowly just to see where the potassium eugenolate first began to melt and then boil. There was no exothermic reaction at 100 deg C this time for obvious reasons. Thus, I'm convinced that the above figures indicating the m.p. and b.p. of the potassium eugenolate are pretty accurate.

Water was added to the flask to try and dissolve the potassium eugenolate. This was the first step taken towards recovering the suspected isoeugenol. And it worked only moderately. Dilute HCl and then concentrated HCl were added; this too worked only moderately. Next,10% H2SO4 was added, working ideally as it neutralized all of the base and potassium eugenolate quicly and easily. The whole mixture was then transferred to a large one-gallon empty and clean mayo jar and diluted with distilled water. The suspected isoeugenol floated to the top (it was opaque and dark brownish-black). It was extracted with trichloroethane (OTC) and worked up using standard techniques. Vacuum distillation of the oil provided isoeugenol as the sole product which through distillation comparisons, has a b.p about eight to ten degrees higher than eugenol.

Catalyst was salvaged as a dirty oil which presumably can be used again. Meager attempts to recrystallize it using petroleum ether failed. This reaction is best repeated with the infinitely cheaper and more readily available PTC catalyst Aliquat 336 (or whatever it's called). Also, it is my belief that magnetic stirring is not a requirement as the boiling phase provides, in my opinion, the amount of agitation required to complete the isomerization. A simple reflux set-up with slow heating up to the boiling poing of the solution (with no temp measurement) for fifteen or so minutes should work ideally. Increasing the reaction time could allow for a reduction in both KOH and PTC catalyst.

Okay, here goes. I search CA all the way from the beginning to about 1962 or something. Basically, I stopped looking when the index indicated that eugenol would from then on be listed as 1-allyl-3-methoxy-4-hyroxybenzene. I didn't look under "isoeugenol" so there might have been a few isomerization abstracts that I might have missed. The ones I'm about to list were located by searching under "eugenol" in the index.

In no particular order:

Conversion of eugenol and its ethers into the corresponding propenyl compounds.
J. Soc. Chem. Ind. 59, 275-6 (1940) [In English]. CA 2485 (1941)

Eugenol (1000g) (I) is mixed in an iron still with 1 L of aqueous solution containing 450g of KOH. H2O is removed at 80-100 deg C @ 20-50 mm of pressure, 300g of diethylene glycol (II) and 100g of triethanolamine (III) are added and the mixture is heated until a thermometer indicates the reaction has occurred (absorption of heat). About 3 minutes at 160 deg C is required. The melt is run into H2O, neutralized with dilute H2SO4 and extracted with benzene. The isoeugenol (largely the trans isomer) is distilled at 1-3 mm yielding 900g of isoeugenol.

Conversion of eugenol to isoeugenol. S.K.
J. of Indian. Inst. Sci. 6, 241-55 (1924). CA 1995 (1924)

Kind of a crappy method. Complete conversion of eugenol into isoeugenol can be brought about by fusing with eleven equivalents of KOH for 5 minutes at 220 deg C or with 8 equivalents for 40 minutes at 220 deg C. Fusion with NaOH at 200-10 deg C produces little or no conversion. The addition of parrafin as a diluent does not effect conversion. In effect, you need either 1100 or 800 grams of KOH to convert 100 grams of eugenol into isoeugenol. Too much base but fast though.

Catalytic reactions. II. Induced reaction. 1. Isomerization of eugenol.
Ricki Horiuchi. J. Chem. Soc. Japan 45, 209-29 (1925); c.f CA 19, 1081. CA 2671 (1926)

This one is really weird. I quote:

Most examples given for induced reactions in the literature are those of induced reaction. According to Nagai's method (C.A. 14, 2839), isomerization of safrole can be easily accomplished by boiling 70% safrole with 20% KOH at 180-200 deg C for 5-6 hours. If eugenol is treated by the same method, practically no isoeugenol is found. If, however, safrole and eugenol were treated together by N's method, both safrole and eugenol were completely isomerized. When safrole was boiled with 15% KOH at 160-170 deg C for 3 hours even safrole is not isomerized at all, yet by addition of and equal amount of eugenol to the safrole, not only isoeugenol is produced, but also the safrole is now completely isomerized. When eugenol alone is used with alkali the mixture becomes very viscous and is difficult to stir but the presence of safrole makes the mixture more fluid. In order to see if the change in viscosity alone might account for the isomerization of eugenol in the presence of safrole, H. dissolved eugenol in benzyl ether (bp 297-8), PhCH2CH2OH (bp 218-20) and isosafrole (bp 253-3) and tested for isomerization. The results show that with PHCH2CH2OH, 65% isoeugenol was obtained, but none of the other two, showing that viscosity alone is not responsible. The reasons why this isomerization is due to mutual induced reactions are given in detail.

CA 1597 (1928)

Vanillin. R.H Bots and Soc. Anon. Produits Chimiques Coverlin. Brit. 271,819, May 25, 1926. The allyl side chain of compounds belonging to the allyl-phenol group is transformed into the propenyl chain, e.g. eugenol is converted into isoeugenol by heating with KOH in the presence of an amino compound such as PhNH2, o-toluidine or o- anisidine. The proceedure may be applied to oil of cloves as an intermediate step in production of vanillin. Cf. CA 21, 3908

CA 2127e Vol. 53

Synthesis of some isomeric methoxyallyldiethylaminoethoxybenzenes.
Zur. Obshchei Khim. 28, 2239-42 (1958).

Edited version:

Heating eugenol with 1:1 aqueous KOH for 3.5 hours at 170 deg C under nitrogen gave 85% isoeugenol, bp @ 8 torr 128-130 deg C.

Here are two more references that use Pd metal and Raney Nickel to effect the isomerization that didn't make my top ten list (too expensive and elaborate):

CA 1002 sec 6 (1937) & CA 927i vol. 46.

So it would seem that to isomerize eugenol to isoeugenol, it's going to take a little more than simple refluxing in alcoholic KOH like the way one does to go from safrole to isosafrole.

Musings on the topic of synthesizing MDMA from Eugenol

MDMA from Eugenol - Someone was asking how this could be done.

What would happen if you heated eugenol with alkali to 200 C? Protocatechuic Acid has been prepared from vanillin by heating vanillin at 240 C with KOH/NaOH. [I A Pearl, Org. Syn. Coll. Vol. 3, p 745] What could you expect if you treated Eugenol to similar conditions? 2-allyl-catechol? If this was methylenated it would give isosafrole.

MDMA from Eugenol

A more likely method. The following method is tedious - but has the virtue of using no lab. chemicals only those obtained OTC from the hardware/pharmacy/grocery shops]. Safrole is not used because I can't get my hands on any. This method could also be used by other Safrole deprived chemists such as those who live in the UK. I haven't yet tried it out. But, it works in theory and if not all processes work in practice then you should be able to substitute another process. Scale-up the processes below as appropriate.

MDMA from Eugenol - Outline.

Eugenol is extracted from cloves or clove oil. Eugenol is refluxed with HCl to give catechol-2-chloro-propane. This is easily converted to the alcohol and then methylenated to give MD-P2Pol. The alcohol is oxidised to ketone which is then reductively aminated with MeNH2 to give MDMA.

Extraction of Eugenol from cloves

see "Experimental Organic Chemistry"; Durst, Gokel, Durst, Gokel; McGrawHill; 1980; p. 467.

50 g of whole cloves (from a supermarket - you can buy cloves in Kg quantities from Indian grocery stores) placed in a 500 ml rb. 3-necked flask with 250 ml of water and several boiling sticks. Steam dist. for 50 to 75 min., with water volume kept constant at 250 ml. Distillate transferred to a separating funnel, extd. with 2 x 50 ml CH2Cl2. Combined CH2Cl2 portion then extd. with 3 x 50 ml 5% KOH soln. (heat is evolved). Combined KOH portion washed with 25 ml CH2Cl2. Aq. layer transferred to a 600 ml beaker and slowly acidified with 5% HCl to a pH = 1 (tested using indicator paper). Aq. layer extd. with 2 x 40 ml CH2Cl2, combined CH2Cl2 portions washed with 25 ml of water followed by 25 ml of half saturated NaCl soln. The CH2Cl2 portion dried over anhydrous granular NaSO4, decanted, CH2Cl2 removed on a steam bath. Pure, 98% Eugenol is obtained as a pale yellow oil. [Scale-up as you see fit].

Phase transfer cleavage of phenolic ether

Using HBr & surfactant, Landini, Montanari, Rolla, Synthesis, 1978, 771

Mixt. 1 mol ArOMe, 560 ml, 5 mol 47% HBr, 50 g, 0.1 mol HPB [hexadecyltributylphosphonium bromide], stirred & refluxed at 115 C, 5 h. Organic layer separated, extracted yd 91% phenol. Distn. residue was dissolved in hexane to recover 46 g, 92%, of pure phosphonium bromide was Mp. 54-56 C. Note MeBr is not recovered as it is a gas at room temp. MeBr is poisonous. Reaction rate is not effected by the nature of the onium salt provided that it is completely soluble in the organic phase. Eg. tetraoctylammonium bromide or trioctylmethylammonium chloride can also be used as catalysts.

Alkyl-aryl Ether Cleavage Using HCl & surfactant
B Jursic, J Chem. Research (S), 1989, 284-5.

Mixt. of 1 mol phenol ether, 50 mol (4 L) HCl (37% aq.) & 0.1 mol (36.4 g) CTAB [cetyltrimethylammonium bromide] stirred under reflux for 36 h. Mixt. diluted with 500 ml water & extd. with ether [DCM is OK substitute here]. Ether ext. dried over MgSO4 & dist. Products purified by distn. Yd: phenol 65%, MeCl is a poisonous gas. Note: HCl is used in concn. 20 to 50 molar excess. 37% aq. HCl gives best results. Surfactant can be recouped. Use of HBr rather than HCl gives higher yields. [Note 2: the chloromethane produced is very volatile and quite poisonous - but it can be dissolved in alcoholic ammonia solution and with react to give methylamine - which is always useful to have].

Hydrogen halide will add across the double bond to give a secondary alkyl halide. This reaction requires a lower temperature than that for the methoxy cleavage. PTC speeds up the addition of both HCl and HBr to allyl benzene compounds. [Addition of hydrohalogen acid to alkenes: Landini & Rolla, JOC 45, 3527, (1980).]

The two operations above (addition of halogen halide and cleavage of the methoxy group) can be combined in one operation using the conditions for the cleavage. Use of HBr will require less time, milder conditions and give higher yields than HCl.

The 2-halo-propan-catechol will be prone to polymerisation under basic conditions. Don't treat this with an alkali or you'll get a gooey mess. You'll have to carefully separate it from the concentrated acid by neutralising acid with bicarb., then extracting. Next, we need to close that catechol with a methylenedioxy bridge. The methylenedioxy-bridge can't be closed at this stage because of that 2-chloro-propane. The alkyl halide has to be converted to something less reactive - an alcohol will do fine.

Convert the catechol-2-halo-propane to an isopropanol group.

[H. A Zahalka, Y. Sasson, Synthesis 1986, 763.].

This is a two-stage process but can be done as a one-pot conversion by first reacting the 2-chloro-propan-catechol with sodium formate and a PTC and then hydrolysing the ester with dilute alkali. The isopropyl-catechol produced is much more stable and can be methylenated.

The process of Methylenation.

[I don't like any of the five methods mentioned in Strike's book]. This 6th method is the best. It can be done entirely with OTC chemicals.


  • Jap patent. 84 046 949-B, To Takasago perfumery KK, 1984. 7 pages in Japanese, abstract available in Jap pat. Abstracts - see appendix 1.
  • Brit. Pat spec. 1518064, Appl. No. 2653/77, Filed 21-1-77; Appl. No. 19735, Filed 30-1-76 in Italy (IT), Complete Spec. published 19-7-78. (To Brichima S.P.A. of Milan, Italy). This is also available in Italian and German. See Appendix 2.
  • Z Yiuguing et al, Jilin Daxue Ziran Kexue Xuebao 2, 92, (1983) [aka Acta Scientiarum Naturalium: in Chinese {any Chinese speakers out there who can translate this please?}
  • "Williamson synthesis of ethers": B Jursic; Tetr. 44, 6677, (1988).]


To your isopropyl-catechol add:

  1. 1.5 mole equivalent of strong alkali [NaOH or KOH]
  2. four mole equivalent of DCM [dichloromethane]
  3. one-tenth mole equivalent of PTC [available from "hair conditioner" or "fabric conditioner" or both]
  4. a trace of iodine [1/100 mole equivalent will do] or an iodide. [available OTC as "Tincture of iodine", or you can get iodine from seaweed (if this sounds tedious - remember that I said no lab. chemicals were needed). The iodine is needed to act as a promoter in this reaction.
  5. Stir vigorously [750 rpm] at reflux for several hours. [The articles above mentioned the use of pressure but I think they use pressure in order to carry out the reaction at a temperature above that of the bp of DCM - the reaction is faster at the higher temperatures].

Isolate and purify the MD-P2-Pol produced - this could be quite tedious. I can think of no other method apart from fractional vacuum distillation.

Oxidise MD-P2-Pol to MD-P2P

This can be done using a reaction analogous to the "cold cat" method or ...

Alcohols can be oxidised to ketones with bleach [R. Stevens et al, J. Org. Chem. 45, 2030, (1980); P. L. Anelli et al, JOC 52, 2559, (1987); J. R. Mohrig et al, J. Chem. Educ. 62, 519, (1985); P. L. Anelli et al, J. Org. Chem. 52, 2559, (1987)]

Oxidation of alcohols to carbonyl cpds. using bleach is quite easy. Liquid bleach at about 5% concn., or higher, is neutralised to pH 8.4. [yes a pH meter is essential - but a pocket $40 job will do]. A PTC in aq. soln. is used to allow the OCl- anion to penetrate into the organic phase. The mole ratio of bleach to alcohol is from 1.05 to 1.1. Swimming pool bleach may also be used. There are some more recent papers that achieve higher yields but require Br- and exotic PTCs. (60 - 85% yield, depending upon the specific PTC and conditions). The problem with using bleach is that the aromatic ring may also be attacked to some extent. [but this is not likely to be a major problem]

The ketone is isolated and purified using the bisulfite addition method. Methylamine [see prep. From Hexamine elsewhere] is added and the Schiff base reduced using available methods [Al amalgam or electrolytic reduction.] to MDMA.