At this point the chemist has what can be loosely considered as pure safrole. She can use this for some methods or can convert it to isosafrole for others. Isosafrole is the runner-up precursor for making X. It cannot be found in nature but rather is made from almost exclusively safrole.
Strike had previously written that about the best recipe for making isosafrole was from boiling safrole with concentrated KOH (Potassium Hydroxide) in ethanol. This was actually the only
recipe Strike knew of at the time. There are numerous basic references for it in the literature and a couple of very convincing examples in the landmark book "PIHKAL" by Alexander and Ann Shulgin. Pioneers in the research of amphetamines they are. Sounds good enough for Strike so someone other than Strike gave it a few shots.
What happens when isomerization is performed is that the double bond of the allylbenzene safrole migrates to the more energetically favorable position between the alpha and beta carbons of the propyl side chain. Thus isosafrole, a propenylbenzene, is bom.
To isomerize safrole to isosafrole one would like to have pure safrole to start with. This, usually, is not the case. Quasi-pure safrole from sassafras oil is ok. Straight-up sassafras oil is probably ok too, though not recommended. The safrole is then refluxed (boiled under a condenser) in a saturated KOH/ethanol solution for about a day and that's it. The temperature of reflux is about 120-140°C owing to the fact that the ethanol (usually boiling around 65-70°C) is saturated with the halide salt.
Now, the ethanol used is almost always anhydrous, meaning it has no water. The closest one can get commercially to anhydrous ethanol is Everclear which is 95% ethanol and 5% water (190 proof). A lot of chemical supply stores will not carry 100% (200 proof) ethanol because it is a potable (drinkable) product. This means that they would have to get a liquor license or some other state permit to sell the stuff and that is a hassle that many don't want to bother with.
The 95% ethanol of Everclear is not an arbitrary concentration that the producer decided to stop at, mind you. It so happens that 95% ethanol and 5% water is a constant boiling mix that no more ethanol can be purified from. That 5% water is there to stay! There are ways to remove that water such as producing a ternary azeotrope by the addition of benzene, but they are kind of a hassle and may
not be necessary. Why? Because an alternative reaction for iso-merization would be to use a saturated aqueous (water, Zak) KOH solution instead of an alcoholic one. This will raise the temperature of reflux to over 200°C but is not too bad an can be used if desired. So, if it is ok to isomerize in pure water, then the little bit of water (say 5%) in Everclear shouldn't effect things too greatly except raise the reflux temperature a tad.
What are some of the alternatives for this procedure? Well, one can use NaOH (lye) in place of KOH but the yields will go down. Also, what about using denatured alcohol instead of pure ethanol? Denatured alcohol is ethanol contaminated with 5-10% methanol. The methanol is there because it is poisonous and prevents people from using the ethanol for drinking. This means that places like the giant hardware stores can carry gallons of cheap contaminated ethanol as many of you have discovered. This product CAN be used.
So now that we have all the reagents out of the way let's see how the reaction proceeds. There's the clear- yellow "safrole" sitting in the bottom if the flask and the clear saturated KOH solution is dumped in. The solution is heated to reflux etc. and yes, some brown byproducts and destruction artifacts will appear. Especially if the safrole is not pure. These byproducts should be expected to some extent because concentrated basic (OH) solutions can be as nasty as concentrated acidic solutions. One is mindful that KOH is less intrusive towards the delicate methylenedioxy ring structure of the safrole/isosafrole molecule.
After 12-24 hours of reflux the reaction is, for the most part, complete. The reaction mix will be a dark brown. So what does one do about all those brown particles and junk. Well, usually there aren't any. The solution should be uniformly dark. If any solids can be seen it means that they are insoluble in ethanol and can be removed from solution by gravity or vacuum filtration through a coffee filter or some paper towels. If it takes a day to drip through the filter then so-be-it. The ethanol with its payload of isosafrole will
still be dark regardless and washing this ethanol with water at this point cannot be done because water is infinitely soluble in ethanol (and visa versa) so two layers will not form.
Filtered or not, the ethanol/isosafrole/(sometimes junk) is relieved of its ethanol by openly boiling it on the stove or by distillation. Any ethanol removed by distillation is saved because it is clean and perfectly reusable (remember, it will still have that 5% water). One should not remove all of the solvent this way because things can get really hot really fast as the last of the ethanol boils away.
Most likely one will end up with a dark oily layer at the bottom of the flask which is perfectly normal. Any solvent that remains will be negligible because in the next step the chemist is going to overwhelm it by adding a buttload of water. If there is about 100-200mL of dark oily layer then a liter or more of water is pored in and mixed really well. After things settle one will see a nice little layer of clear isosafrole form at the bottom, a murky brown lyikesl emulsion layer in the middle and a blackish water layer on top. The oil is separated and the emulsion is dealt with in any way that is best from the below mentioned strategies for emulsion fighting. This particular emulsion is of very fine particulates and has a lot of isosafrole in it. If one is patient then one can add solvent (DCM is good) to the remaining water/emulsion layer, shake then vacuum filter. The crud remaining in the filter is the junk that was causing all the trouble. With it gone one can now extract the remaining water layer a couple more times with solvent, add the solvent to the isosafrole oil then remove the solvent by boiling or distillation to get pure isosafrole ( 70-80% trans, 20-30% cis).
NOTE: a really good procedure for busting up the emulsion caused by introducing the water is to slowly acidify the water layer with HCI. As the water layer acidifies one will see the emulsion vanish. As this occurs the water layer will lighten as particulates and stuff start to exit the water layer and go into the DCM layer. Works every time.
If one is still having trouble then there is the old failsafe: just throw the whole mess into a distillation set up and fractionally distill. Solvent comes first, water second and (eventually) isosafrole will come as well.
Sound easy? Well, procedurally it is. But the yield can really suck sometimes and it can get really messy. Lots of people have not been pleased with it. Strike wasn't. But what can you do?
Well, someone named Osmium sent Strike a journal reference about doing the exact procedure above but carrying it out in a microwave ! A group of Brazilian folks tried both conventional and microwave heating. They had strangely high yields (all above 90%) using both ways. Both 2M and 4M KOH concentrations were used in ethanol, propanol and butanol. They also found that adding 2M KCI along with the KOH helped things. The really cool things were that when they used butanol and 4M KOH they only had to apply 15min of boiling to get a yield of 99%! Still the other reactions in propanol and ethanol only needed 2-5hrs for 90+% conversion. Strike would assume that the reduced reaction times afforded by the larger alcohols and such may very well help reduce tar and destruction and make for a better procedure.
Using the microwave just decreased the reaction time to 3-30minutes. The dudes in the article used a household, 500W Brazilian microwave (Yikes!). They cut a whole in the top of the microwave to allow the condenser apparatus to pass through the oven. They then killed themselves most likely. But not before they were able to scratch down this procedure as they slowly burned to death:
"GENERAL PROCEDURE: The Alcoholic alkaline solution is prepared by prolonged stirring of 8.8g (or 4.4g) of KOH pellets in 30mL of alcohol. The alkaline solution is placed in a round-bottom flask provided with a reflux condenser (microwave or conventional systems). Then 4.0g of Safrole (or eugenol) is added and the solution heated."
Gee, that all sounds fine and dandy. But Strike saved the best for last. Without a doubt, the cleanest, fastest and easiest method for isomerizing safrole and other allylbenzenes is to use CaOH. Someone who is 'Not Tim' (Strike's name for her) emailed Strike the Chemical Abstracts entry for the procedure:
"CA 47:9360. Isosafrole from safrole. Yoshiharu Ogata (Regeneration Camphor Co.). Japan. 5331 ('51), Sept. 15. Safrole, b. 132-4°, 100, CaO 15, and KOH 1 g. are heated 15 min. at 243.5°, the CaO and KOH filtered off and the filtrate distd. to obtain 95% isosafrole, b. 240-8°. K. Kitsuta"
'Not-Tim' said Strike was wasting everybody's time with the etha-nol/KOH recipe and you wanna know what? He was kinda right (But, those Brazilian improvements above aren't here just to take up space!). This method is sooooo easy!
All one does is follow that CA recipe to the letter. In a small flat-bottom flask place 100g safrole, 15g CaO and 1g KOH. Now the flask is heated directly on a stirring hotplate (stirring is nice but not necessary). If one has a round-bottom flask or a low output heating source, a shallow oil bath can be employed to get even heating. The really neat thing about this method is the proof-positive heat gradations that will occur. The mixture will rise to the boiling point of safrole (232-234°C). And it will hold there for just a brief second. Then it will start to bubble and the temperature will rise! Sure enough, after only 15min of this the temperature will reach the near boiling point of isosafrole (~243-250°C). And that, as they say, is it.
The oil that remains is only slightly dark. Definitely translucent which is always lovely. One can then add some water to take up the unreacted salts and separate the oil from this. If an emulsion forms it can be busted up with some 10% HCI. That isosafrole is clean enough to proceed with but it can of course be distilled for ultra purity.
All of the above isomerization recipes can, in theory, produce almost 100% yields of isosafrole. But about 20-30% of that isosa-frole is in a screwed up configuration called 'cis'. This cis isomer does not react the same way as trans and the drug that will be made from it will not be recognized in the same way as trans in the brain cells of users. Tsk.
The chemist can try to separate the two isomers by careful fractional distillation but it will be next to impossible to do because both the cis and the trans have practically the same boiling point. There are a few things that the chemist can do or hope for to get rid of that cis isomer. The cis configuration is less stable than the trans and may switch over to the trans configuration with a little help. The chemist can gently heat the isosafrole oil to about 150°C for an hour or so. She can also try the same heating except have the oil mixed with some acetic acid. Also, the isosafrole can be fractionally distilled to ultra purity and then be allowed to simply sit for a couple of days. Trans isosafrole 'may' spontaneously crystalize out while the cis form remains as an oil. They can then be separated by filtration.
When all is said and done, the chemist may just wish to leave the damn cis isomer in with the trans. It's not really going to hurt anything and if it goes unreacted in some of the conversion steps it will be lost in the process long before X is made. It might even correct itself during some of the conversion steps.
Continue reading here: Other Precursors
Was this article helpful?