Join Date: Feb 2007
Like this post?
Studies On The Production Of TMA-2
That route has several drawbacks which make it impractical for clandestine synthesis. The first and most important problem is the availability of 2,4,5-trimethoxybenzaldehyde. This substance is not exactly a linchpin of chemical commerce. So far as I know, it has one use: making TMA-2. Those same folks who gave me the hassle over the purchase of Rochelle salts will certainly report a shipment of 2,4,5-trimethoxybenzaldehyde, and the heat will not be far behind.
Further chemical supply problems arise from this method's use of large amounts of anhydrous ether or THF in the LiAlHj reduction. This too will be duly noted by the heat, especially in combination with buying LiAlHt. A much more low-profile synthetic route is possible using calamus oil as the raw material. A couple of patents granted in the late 80s have completely changed the field of psychedelic amphetamine manufacture from the way Dr. Shulgin knew it during his days of cooking in the 60s. Previous to the publication of these patents, the Knoevenagel condensation of benzaldehydes to yield the nitroalkene, followed by the reduction of the nitroalkene to the amphetamine, was far superior to an alternative route making use of the common essential oils.
Many essential oils have as major components substituted allylbenzenes. For example, sassafras oil is 80-90% safrole: The alternative route was to take this substituted allylbenzene, move the double bond to the propenyl position by heating with anhydrous alcoholic KOH, yielding in the case of safrole, isosafrole. Then a messy, tedious and low-yield reaction was used to convert this propenylbenzene to the corresponding phenylacetone. All we veteran speed cooks love phenylacetones, because they offer the cleanest and best route to the amphetamines, but the old-fashioned method of converting propenylbenzenes to phenylacetones made this route impractical.
My own experience with this reaction dates to the early 80s, when I decided to torment myself by trying it. Detailed cooking procedures using it can be found in Pikhal under MDMA. My experience with the KOH isomerization was that the conversion of safrole to isosafrole went cleanly at about 100% yield, as long as traces of moisture were excluded from the reaction. The conversion of isosafrole to methylenedioxy-phenylacetone is another matter. The yields are low, a lot of work is required because the formic acid and hydrogen peroxide must be removed from the reaction mixture under a vacuum before final treatment with sulfuric acid solution to yield the phenylacetone, and these vapors corrode the aspirator supplying the vacuum. This method stinks!
Two patents dating to the late 80s, and to a lesser extent a journal article dating back to 1970, have turned the situation around. The first patent I will cite is US patent 4,638,094, titled "A Process for Producing Phenylacetones." This patent reveals, using many different examples over the course of 36 pages, the best general method for converting allylbenzenes to the corresponding phenylacetone in very high yields.
This procedure reacts the allylbenzene (for example safrole, as obtained in pure form by vacuum distilling sassafras oil) with methylnitrite in methanol solution containing water and a palladium catalyst to yield the phenylacetone. The palladium catalyst can be used in a variety of forms, as detailed in the patent. The best choices for use with safrole are palladium bromide, chloride, or a mixture of palladium chloride and copper chloride. Of the three, the mixture catalyst is better for reasons which will be explained in the following cooking example:
In a 4000 ml beaker, or one-gallon glass jug, is placed 3000 ml methyl alcohol, 150 ml safrole, 300 ml distilled water, and the chemist's choice of either 20 grams palladium bromide or ten grams of palladium chloride or a mixture of one gram palladium chloride and 4.25 grams copper chloride (CuCk). The catalyst choices have been given here in order of good to best. The reason why the last choice is best is because of the very high cost of palladium salts. Palladium chloride is preferred over the bromide because palladium chloride finds use in the electroplating field. It is used there in baths to plate palladium, and as part of the activation process to prepare plastics to be plated. The bromide is not as commonly used. Next, a methyl nitrite generator is rigged up as shown in Figure 3:
Into the 2000 ml flask is placed one pound of sodium nitrite, 225 ml of methyl alcohol, and 260 ml of water. They should be swirled around for a while to
mix. Then 680 ml of cold dilute sulfuric acid (made by adding 225 ml of sulfuric acid to 455 ml of distilled water, mixing and chill-ng) is put into the dropping funnel.
Now vigorous /V.2000 ml magnetic stirring is begun in the beaker or glass jug containing the Figure3 allylbenzene-alcohol-pal-Methyl nitrite generator. In the 1-mole batch given in this example, about 6 moles of methyl nitrite are bubbled into the reaction mixture, while only 2 are required for the reaction. The reason for the excess is because methyl nitrite is not held in solution very well on account of its very low boiling point. If ethyl nitrite was used instead, then only three or four moles would be needed.
While the reaction is being done, the mixture takes on the appearance of mud if palladium bromide is being used. A fizzing also occurs, which gives the reaction mixture the appearance of freshly poured Coke. Note above that a bit of acid is required to get hydrolysis of the intermediate dialkoxyphenylpropane to the phenylacetone.
The best pH for this reaction is between 4-7. If palladium chloride or the mixed catalyst PdCh-CuCla is being used, the pH of the reaction mixture can be adjusted to this range by adding a small amount of HC1. If PdBr2, is used, it is best to wait until the catalyst is filtered out before adding HC1, as the HC1 could form PdCh and complicate catalyst recovery. The pH of the reaction mixture is best measured by first dampening some indicating pH paper with distilled water, then putting a drop of reaction mixture on the paper. The preferred temperature for this reaction is about 25° C throughout.
When all the methyl nitrite has been bubbled into the reaction mixture, stirring should be continued for another hour. Then, if palladium bromide was used, it should be filtered out. Repeated filtrations will be needed to remove all of the catalyst, because it gets quite finely divided during the course of the reaction. This leaves a clear light-reddish solution. If palladium bromide was used, now adjust pH to 4-7, and allow another hour to complete the hydrolysis.
If palladium chloride or the mixed catalyst was used, these substances are soluble in alcohol. In this case, the catalyst will be recovered later. Here, check the pH of the solution again to be sure it is in the proper range before proceeding.
Now the alcohol solvent must be removed. This is best done by pouring the reaction mixture into a large filtering flask, stoppering the top of the flask, and removing the solvent under a vacuum. Use of a hot-water bath to speed evaporation is highly recommended for this process. It is not OK to distill off the alcohol at normal pressure, as the heat will cause the nitrite and NO in solution to do bad things to the product.
To the residue left in the flask after removal of the alcohol, add some toluene to rinse the product out of the flask into a sep funnel. Next, put 300 ml of water into the flask to dissolve the catalyst if PdCla or the mixed catalyst was used. Add the water solution to the sep funnel to dissolve carried-over catalyst there, then drain this water solution of catalyst into a dark bottle and store in the dark until the next batch. If PdBr2 was used, this step can be skipped. Just store the filtered-out PdBra under water in the dark.
Now the toluene-phenylacetone solution should be distilled through a Claisen adapter packed with some pieces of broken glass to effect fractionation. The first of the toluene should be distilled at normal pressure to remove water from solution azeotropically. The b.p. of the azeotrope is 85° C, while water-free toluene boils at 110° C.
When the water is removed from solution, turn off the heat on the distillation, and carefully apply a vacuum to remove the remainder of the toluene. Then with the vacuum still on, resume heating the flask, and collect the substituted phenylacetone. Methylenedioxyphenylacetone distills at about 140° C and 160° C using a good aspirator with cold water. A poor vacuum source leads to much higher distillation temps and tar formation in the distilling flask. The yield from the reaction is close to 150 ml of phenylacetone. Its color should be clear to a light yellow. The odor of methylenedioxyphenylacetone is much like regular phenylacetone, with a trace of the candy shop odor of the safrole from which it was made.
A higher-boiling phenylacetone like 2,4,5-trimethyloxyphenylacetone is better purified as the bisulfite addition product, unless a vacuum pump giving high vacuum is available. To make the bisulfite addition product, take the residue from the filtering flask, dissolved in some toluene and freed from catalyst as described above, and pour it in a beaker. Next, add 3 volumes of sodium bisulfite solution prepared by adding sodium bisulfite or metabisulfite to water until no more dissolves. Shake or vigorously stir for a couple of hours to convert the phenylacetone to the solid bisulfite addition product. Filter out the solid, then regenerate pure phenylacetone by putting the solid into a round-bottom flask, adding an excess of saturated solution of sodium
bicarbonate in water, and refluxing for a couple hours. After cooling, the phenylacetone should be extracted out with some toluene. The toluene should then be removed under a vacuum, and the residue stored in a freezer until conversion to the amphetamine. All phenylacetones are sensitive to light, and should be stored in the freezer.
The above cooking procedure is the best way to process allylbenzenes to the corresponding phenylacetones. Sassafras oil, as previously mentioned, is 80-90% safrole. Calumus oil, if its country of origin is India, consists of about 80% of the allyl isomer of asarone:
It too can be purified by distillation under a vacuum to yield fairly pure allyl-asarone. Its boiling point is 296° C at normal pressure and about 170° C with aspirator vacuum. More details on this Indian calamus oil can be found in Chetn. Abstracts column 6585 (1935), also Current Science, Volume 3, page 552 (1935).
My search for calamus oil of Indian origin came up empty. In fact, the health-food store in my town, which is well-stocked with various oils for use in aromatherapy, had never heard of the stuff, nor was it listed for sale in their catalogs. This left one alternative: dig up the roots of North American calamus, and steam-distill the oil out of them.
While searching for calamus in my area's swamps, bogs and ponds, the damaging effects of the spread of purple loosestrife was obvious. This imported plant from Europe has taken over much of the former habitat of the calamus plant. Here in America, the loosestrife is free from the insect that keeps it under control in Europe by feeding on its seeds. The state paper-pushers have been thinking for years about importing the bug, without ever getting off their butts and doing it.
I suggest this project to somebody out there in the reading public so that
it can finally get done while there is still some native flora left. After a lot of searching, I finally found a large patch of the American calamus. (See Figure 4.) The time for harvesting the roots of the calamus is in the fall after the killing frost. The frost brings the oil down out of the leaves and
into the root for winter storage. The roots are about a foot long, an inch or so in diameter, and run horizontally in the soil at a depth of a few inches. They are best dug out using a fork, taking care not to pierce the root, as this will cause loss of oil during drying. The dug-up roots should be rinsed free of dirt, and the tops cut off there in the field. (See Fig 5.)
The roots should then be taken home and allowed to dry at room temperature for a week or two. Take care that they do not get moldy!
Once dried, oil can be distilled from them. This is done by first grinding up the roots in a blender or with a Salad Shooter, and piling the ground-up roots into a large pressure cooker. A good-sized pressure cooker will take a load Of 10-15 pounds Of Calamus plant root and fibrous rootlets.
Next, add a few gallons of water, a couple handfuls of salt, and mix. The oil can now be distilled. Attach a five-foot length of copper tubing to the steam exit on the lid of the pressure cooker. Its diameter should match that of the steam exit so that steam is not lost here, and should be tightened into place with a pipe clamp. The tubing should then be led downward into a pail of ice water, and back up into a dark-glass 40 or 64 ounce beer bottle. The ice water cools the steam, turning it into water which collects in the bottles.
Heat is applied to the pressure cooker, bringing it to a boil. Heat as fast as is possible without bringing over foam or having uncondensed steam escape. When a couple of gallons have been distilled out, stop the heating and add a couple more gallons of water to the pressure cooker. Continue this process until 4-5 gallons of water have been collected.
This process is a steam distillation, and is the way most plant oils are obtained. The steam distillate in the beer bottles contains calamus oil
floating on top of the water and clinging to the glass. Calamus oil produced from American plants is reddish brown, and has a strange,
pleasant and sweet odor. For more detailed information on calamus oil see The Chemergic Digest August 30, 1943, pages 138-40, and Soap, Perfumery and Cosmetics August 1939, pages 685-88.
The oil is obtained by first saturating the steam distillate with salt, then extracting the oil with toluene (obtained off the shelf in the hardware store's paint section). About a gallon of toluene is plenty to effect the extraction. Then the toluene is removed by vacuum evaporation in a large filtering flask to yield the calamus oil as a residue in the filtering flask after the toluene has been evaporated. The yield is about 200 ml from 15 pounds of roots.
Calamus oil obtained from sources other than India differs from the Indian oil in two important respects. The amount of asarone in the oil is much lower than the 80% found in the Indian oil, and the position of the double bond is propenyl rather than allyl:
The asarone is obtained in pure form from the oil by fractional distillation under a vacuum. Asarone boils at about 170° C under good aspirator vacuum of 15-20 torr. The asarone fraction should be collected over a 20-degree range centered on 170° C. I found the yield of asarone from American plants to be about 15% of the oil, giving 30 ml from 15 pounds of root.
Asarone is a light-sensitive material, and as such, should be stored in
the fridge or freezer. Upon standing in the fridge, it will crystallize, allowing further purification by filtering. The m.p. of the pure substance is 67° C. Asarone is listed as a cancer-suspect chemical, along with half the other substances in the world. In reality it is not particularly harmful. See Chem. Abstracts 1931, page 169. It also doesn't have any pronounced drug effect at reasonable oral dosage. See Dr. Shulgin's comments on the substance in Pikhal.
With the double bond in the propenyl position, we come to the next major advance over the disappointing procedure cited in the beginning of this chapter. See European Patent 0,247,526 titled "A Process for 3,4-dimethoxyphenylacetone Preparation." This process uses a simple electrochemical cell to convert the propenyl-benzene to the corresponding phenylacetone in very high yield. The procedure given also works with 2,4,5-trimethoxypropenylbenzene (asarone), and probably also with iso-safrole. It is my opinion that it will work with all propenylbenzenes.
There are great advantages to the use of an electrochemical cell in
clandestine synthesis. The solvents and the salts can be reused over and over again, making for a very low profile. The reagent doing the transformation is electricity, available at the nearest wall socket. The transformer, multimeter and alligator-clip wiring can all be obtained at Radio Shack with zero suspicion attached. This method comes with
my highest recommendation.
To do the reaction, a 1000 ml beaker must be rigged up as shown in Figures 6 and 7. A central piece of stainless steel having a surface area of about 100 cm2 actually in contact with theGraphite anodes (2) solution is securely clamped into place down the center of the beaker. On each side of this stainless steel piece, securely clamp into place two pieces of
graphite, roughly equal in size, having a total surface area in contact with the solution of about 70 cm2. All three of these electrodes should run
straight down into the flask, and a constant distance of 1 cm should
separate the surface of the anodes from the Electrochemical cell used to convert a cathode. This is verypropenylbenzene to the corresponding phenylacetone.
important, as the anode-to-cathode distance determines the voltage at which this cell runs. It is also very important that shorts between the anode and cathode be prevented. The current must flow anode-to-cathode through the solution, not through a short! Then into the beaker place a magnetic stirring bar, 25 grams of NaBr dissolved in 100 ml of water, 500 ml of acetonitrile, and 20 grams of asarone. Note now the depth of the solution in the flask, and be sure that the required amount of electrode surfaces are in the solution. I depicted graphite sheet anodes, in Figures 6 and 7, but the more commonly available graphite rods will work as well.
Now, using alligator-clip wiring, attach one clip to the central stainless steel cathode, and run it to your DC transformer where it is connected to the black or negative pole. Another approximately one-foot long section of alligator-clip wiring is attached to each of the Stainless steel cathode
graphite anodes; i.e. the alligator-clip on one end gets attached to graphite anode A, while the alligator-clip on the opposite end of the wire
gets attached to graphite anode B. Then remove some insulation in the
center of the wire, and make an electric connection to the positive and red pole on the DC transformer.
Next, begin vigorous magnetic stirring of the solution, turn on the transformer, and adjust the output of the transformer so that it is pushing a constant current of about 3.4 amps. All three of the electrodes should be fizzing away at this point. If one appears dead, dig the alligator-clip into it to make better contact. Continue passing electricity until 24,000 coulombs have been passed through the solution. A coulomb is defined as 1 amp-second, so this takes about 2 hours at 3.4 amps. The patent states that the temperature must be kept in the range of 10-306 C, so watch to make sure that the current doesn't heat up the solution too much. Surround the beaker with ice if this occurs.
The electrochemical cell makes the following compound, an epoxide. When the required amount of current has been passed, turn off the juice and the stirring, and pour the contents of the beaker into a sep funnel. Allow it to stand for about Vi hour for the phases to fully separate. An aqueous phase settles out at the bottom of the sep funnel, in spite of the fact that water and acetronitrile are miscible. This water phase contains the NaBr. It should be separated off and saved for reuse.The acetonitrile phase contains the product. It should be poured into a distilling flask, and the solvent removed under a vacuum. By packing the receiving flask in dry ice during this process, the acetonitrile can be recovered for reuse.
The residue of epoxide product left in the flask should be diluted with 150 ml of ethyl acetate, and poured into a 500 ml flask. Flush the flask with nitrogen, then add 1.5 grams lithium iodide, and reflux for 5 hours. The lithium iodide catalytically transforms the epoxide to the phenylacetone.
After the 5 hours of reflux are over, allow the mixture to cool, then pour it into a sep funnel. Wash the ethyl acetate solution with 50 ml of water to recover the lithium iodide into the water solution. Separate off the water layer, and evaporate the water to recover the lithium iodide for reuse. The ethyl acetate solution should be dried over some anhydrous sodium sulfate, then the ethyl acetate evaporated off to give about 20 grams of 2,4,5-trimethoxyphenlyacetone. This light-sensitive substance should be stored in the freezer.
Acetonitrile is a quite poisonous solvent, dangerous both in inhalation from the fizzing electrochemical cell and by absorption through the skin. It has been my experience that just spilling a little bit of it on your skin is enough to give you head rushes and make you feel uncomfortable. The use of acetonitrile can be avoided without loss of yield by using the alternative procedure in Example 6 in the patent.
The electrochemical cell is constructed in exactly the same way as in the first method. Then into the electrochemical cell put 400 ml of dimethylformamide, 200 ml of water containing 27 grams NaBr, and 20 grams asarone. Check the level of the solution, and make sure that the amount of electrode surfaces are the same as in the first method. Then begin stirring, and pass the current through the solution exactly as in the first method.
When the 24,000 coulombs have been passed, pour the contents of the beaker into a sep funnel, dilute with 1000 ml of a 20% solution of salt in water, and extract four times with 200 ml portions of ethyl acetate. The combined extracts, amounting to 800 ml, should be washed twice with 200 ml portions of a 20% solution of salt in water.
The ethyl acetate solution contains the product epoxide. It should be evaporated under a vacuum to a volume of about 200 ml, then reacted with lithium iodide just as in the first method to yield about 20 grams of 2,4,5-trimethoxyphenylacetone.
Recycling of solvents is possible with this method too. Ethyl acetate can be recovered during the vacuum evaporation by use of a dry-ice trap. The dimethyl-formamide can be recovered by vacuum distillation.
The Journal Method
A very effective alternative method exists for converting propenyl benzenes to phenylacetones. I know through mail received from the reading public that this method gives a yield of about 80% when used with isosafrole. Similar results can be expected when used with asarone.
In spite of the high yields and simplicity of this reaction, I can't recommend its use. That's because this procedure uses thallium(III) nitrate to oxidize the propenylbenzene to the corresponding phenylacetone. The thallium(III) nitrate gets reduced to thallium(I) nitrate. Both of these heavy-metal compounds are very poisonous and, unlike organic chemicals, the heavy metals persist forever in the environment, and accumulate in the body. You want a bunch of thallium around the house about like you want to be kicked in the teeth with a heavy pair of boots.
A further bad aspect of this method is its high cost. 100 grams sell for $150, and the high molecular weight of the compound means that a lot of it has to be used to get a moderate amount of product. One pound of thallium(ni) nitrate is required for a 1-molar batch.
This method can be found in Tetrahedron Letters No. 60, pages 5275-80 (1970). To produce a one mole batch, dissolve one mole of propenylbenzene in some methanol, and put it into a one-gallon glass jug. In a beaker, dissolve one mole (448 grams) of thallium(HI) nitrate trihydrate in methanol. Then pour the thallium solution into the jug with the propenylbenzene, and stir at room temperature for 5 minutes.
The thallium(I) nitrate formed by the reaction comes out of solution. It is
removed by filtration. The propenylbenzene has at this point been converted to a ketal. This is hydrolyzed to the phenylacetone by shaking the filtrate with about 2000 ml of 1 molar sulfuric acid solution in water for about 5 minutes. The phenylacetone is then extracted out with a couple of portions of tolulene. This extract is then washed with 5% NaOH solution, then distilled or purified by conversion to the bisulfite addition product.
Production of TMA-2, MDA, etc. from the Corresponding Phenylacetone
There are three good methods for converting the phenylacetone to the psychedelic amphetamine. Choice number one is to use reductive amination with a hydrogenation bomb with Raney nickel, ammonia and alcohol solvent. See Journal of the American Chemical Society, Volume 70, pages 12811-12 (1948). Also see Chem. Abstracts from 1954, column 2097. This gives a yield of about 80% if plenty of Raney nickel is used. The preferred conditions for use with MDA is a temperature of 80 C, and a hydrogen pressure of 50 atmospheres.
The drawback to this method is the need for a shaker device for the bomb, and also a heater. The use of platinum as the catalyst in the bomb works great when making MDMA, but gives lousy results when making MDA. There may be a way around this, however, for serious experimenters. It has been found in experiments with phenylacetone that a mixture of ammonia and ammonium chloride produces good yields of amphetamine (50%) when used in a bomb with platinum catalyst. Methylenedioxyphenylacetone is quite likely to behave similarly, along with other phenylacetones.
To use this variation, the following materials are placed in the 1.5 liter champagne bottle hydrogenation device described in Chapter 11 of Secrets of Methamphetamine Manufacture, Third Edition.
.5 gram platinum in 20 ml distilled water. If this platinum is in the form of
PtO2 instead of reduced platinum metal catalyst obtained with borohydride, the experimenter must now reduce the platinum by pressurizing the bottle with hydrogen and stirring for about an hour.
Next 100 ml of methylenedioxyphenylacetone is added along with 40 grams NHUCl, 500 ml methyl alcohol saturated with ammonia gas, and 50 ml NHjOH. The bottle is then set up as seen in Figure 17 in Secrets of Methamphetamine Manufacture, Third Edition. The hydrogenation is done as described in that section.
When the reduction is over, the contents of the flask are filtered to remove the platinum metal for reuse. Some crystals of NH4C1 are also filtered out; they are rinsed down with some water to remove them. Next the filtered batch is poured into a 1000 ml round-bottom flask, a few boiling chips are added, and the glassware is set up for refluxing. Plastic tubing is attached to the top of the condenser and led outside. The mixture is boiled under reflux for one hour to force out the excess ammonia.
Next, the solution is allowed to cool, and made acid to congo red (about pH 3) with hydrochloric acid. Now the glassware is set up as shown in Figure 3 of Secrets of Methamphetamine Manufacture, Third Edition, and the solution is evaporated to about one-half its original volume under vacuum. A fair amount of crystalline material forms during the acidification and vacuum evaporation.
Next, 400 ml of water is added to the solution, and then it is extracted with about 100 ml of toluene. The toluene layer is thrown away because it contains garbage. The batch is now made strongly basic by adding lye water to it. It should be remembered here that it is very important to shake the batch well once it has been basified, to make sure that the MDA hydrochloride gets neutralized. Finally, the MDA is extracted out with a few hundred ml of toluene, and distilled under vacuum. The boiling point is about 160fi C under aspirator vacuum. The yield is about 50 ml.
Another very good choice of a method for converting methylenedioxyphenylacetone to MDA is the Leuckardt reaction. In this case formamide is used instead of N-methyl formamide. When used with phenylacetone to make amphetamine, only the very high-grade 99% material will work. In the case of methylenedioxyphenylacetone, however, the much more commonly available 98% formamide works just fine. See Chem. Abstracts from 1952, column 11246, and Austrian patent 174,057. In this variation, 40 ml of methylenedioxyphenylacetone is mixed with 100 ml of freshly vacuum-distilled formamide, 2 ml glacial acetic acid, and 20 ml water.
This mixture is heated up to about 130° C, at which point bubbling should begin, then the temperature is slowly raised to keep the bubbling going, as described in Chapter 5 of Secrets of Methamphetamine Manufacture, Third Edition, until a temperature of ISO° C is reached. This should take at least 5 hours. The yield is 70%.
Processing is then done just as in the case of meth. The formamide is destroyed by boiling with lye solution. In this case, the ammonia gas produced is led away in plastic tubing. The formyl amide is then separated, and hydrolyzed by refluxing in a mixture of 60 grams KOH, 200 ml alcohol, and 50 ml water for an hour. After the reflux, the mixture is made acid with HC1, and the alcohol evaporated away under a vacuum. The residue is then diluted with water, and the freebase obtained by making the solution strongly alkaline to litmus by adding lye solution. The freebase is then
extracted out with some toluene, and distilled.
This procedure is no doubt applicable to all phenylacetones. In the case of 2,4,5-trimethylphenylacetone, I would first try this with only half as much added water. Those phenylacetones containing the methylenedioxy grouping, I would use just as stated.
The last choice is a very simple, but also very time-consuming (several days!) reaction. Sodium cyanoborohydride in methanol with ammonium acetate and methylenedioxyphenylacetone at pH 6 react to give disappointing yields of MDA. See Pikhal by Dr. Shulgin in the section under MDA for full cooking instructions.
This method is general for all phenylacetones, as Dr. Shulgin used it on quite a variety of them, all with similar low yields. In all of these methods, once the freebase is obtained in pure form by distillation (the boiling point of the amphetamine is similar to the phenylacetone), the freebase should be converted to the crystalline hydrochloride derivative. This is done by dissolving about 50 ml of freebase in about 400 ml ether or toluene, then bubbling dry HC1 gas through the solution, and filtering out the crystals to dry. See Chapter 5 of Secrets of Methamphetamine Manufacture, Third Edition for a full description.