[www.rhodium.ws] [] [Chemistry Archive]

Oxidation of propenylbenzenes to P2P's using Oxone

by Chromic

HTML by Rhodium

Oxone Technical Information
Discussion of oxone epoxidations

This method is 100% OTC. It is awesome for a first time chemist not interested in yields or time constraints, but rather in producing an active product without the purchase of suspect chemicals.

Running the reaction

Setup a 1L or 2L flask for reflux, optionally monitoring the temperature.

  1. Add 100mmol alkene (eg 16.2g isosafrole, see note)
  2. Add 200mL MeOH
  3. Drop in a 2" stir-bar and start the stirrer.
  4. Add 500 mL distilled H2O.
  5. Add 200 mmol 2 KHSO5·KHSO4·K2SO4 (~72.7g Oxone)
  6. Slowly add 290 mmol (24.4g) NaHCO3

This will form a swirling white brew of chemicals in your flask. Let this reaction run 5h with no external heating.

Extracting the goods

Turn off the stirrer, and let the mixture settle. The sediment will drop to the bottom. Pour off as much as possible into a beaker without getting any solids in. Suction filter the rest. The crystals will be off-white. A lot of your product will remain in the crystals. Rinse the crystals with fresh methanol to try and get as much product out as possible. The crystals should be bright white after rinsing twice.

Goiterjoe's Voice: When you wash the residual salt with methanol and add these extracts back to the mother liquid, more salt will fall out of solution. It helps to flood the liquid with some more water to get these salts back into solution. This also helps you extract more of the yellow color(presumably the epoxide) out of the liquid with your DCM washes.

Load into your 1 or 2L sep funnel and extract 3x100ml (or 4x60ml) with DCM.

Recover the DCM by simple distillation on a water bath.

React the oil as usual, to get ketone. If you don't know how to do the usual reaction, read the small section at the end. Purify the ketone by vacuum distillation and/or forming the bisulfite adduct, then the ketone plug into Methyl Man or Osmium's reductive amination... and away you fly!

Pool chemicals rock. The war is over.


The temperature will quickly rise to ~35°C after mixing the chemicals. The color will change from a swirling mix of white solids in a colorless and clear liquid to white solids in a light yellow liquid. As the reaction comes to completion, the temperature will drop to room temperature.

If using anethole:

  1. The glycol oil will be a clear yellow (instead of clear and colorless, as with anethole) and no longer smell of licorice, rather like root beer... and it has maybe incorporated the smell of burnt rubber.
  2. The ketone will have a stronger non-licorice smell-similar to, but stronger than the glycol/epoxide...
    Less similar to root beer, and more medicinal and smelling like burnt rubber/fruity. Your phenylacetone will be a light-brown oil!

If using isosafrole:

  1. The glycol oil will be a darker yellow (instead of clear and colorless, as with isosafrole) and will smell more like root beer after this... the same sort of burnt rubber smell is present. The yellow color won't be totally extracted by the DCM.
  2. The ketone will have a stronger smell, losing the hints of root beer, and will have that medicinal burnt flower smell. Your phenylacetone will come out dark-yellow. In my experience, all phenylacetones have a similar smell, but each is subtly different. Some yellow color won't be able to be extracted by the DCM when extracting the ketone. The ketone is a dark red before NaOH washings which really clear up the color.


  1. 1.5g anethole (10mmol), the crude yield of ketone after steam distillation was 0.6g (3.7mmol), 37% yield.
  2. 3.0g anethole (20mmol), the crude yield of ketone was 2.2g (13.4mmol), 66% yield
  3. 14.8g anethole (100mmol), the first step crude yield was 16.4g epoxide (100mmol), 100% yield, and the crude yield after hydrolysis and washings was 12.6g ketone (77mmol), 77% yield.
  4. 24.3g isosafrole (150mmol), the first step crude yield was 22.4g epoxide (125mmol), 83% yield (sloppy technique), and the crude yield after hydrolysis, washings and vacuum distillation (~60mmHg, 190-92°C) was 13.4g ketone (75mmol), 50% yield.
  5. 32.4g isosafrole (200mmol) (300ml MeOH, 750ml H2O, 45g NaHCO3, 145.4g Oxone), from which the first step crude yield was 30.7g epoxide (171mmol), 86% yield.

Qualitative Analysis

The product from a Methyl Man or Osmium reductive amination is unmistakably active. Recrystallisation from IPA, followed by a rinse with acetone yields a bright white odorless crystal that is soluble in warm alcohol, insoluble in acetone that is not hygroscopic and completely melts below 200°C. It tests purplish-greenish black with Marquis reagent (liberating hydrogen chloride gas), and burns cleanly. 150mg will produce a smooth high complete with empathy, eye jitters, jaw clenching and pupil dilation. TLC, MS, IR or NMR has not been done for confirmation on the product.


  1. 0.1mol (100mmol) alkene is 14.8g anethole or 16.2g isosafrole. 0.1mol (100mmol) epoxide is 1.6g heavier than each alkene.
  2. 1991 ref said 1g Dupont Oxone contains 2.75mmol KHSO5, although 1985 ref said 0.2mol is 61.5g Oxone (ie 1g contains 3.25mmol KHSO5). I went with the info from the 1991 ref... who really cares anyway, as the reagent is used in excess.
  3. The whole idea of buffering is to keep the pH only weakly acidic... note that the molecular formula for Oxone is 2 KHSO5·KHSO4·K2SO4, so there's 0.3 mol of acidic hydrogen for 0.2 mol of KHSO5... which means we'd need less than 0.3mol of bicarbonate to leave this slightly acidic. Luckily, if too much bicarbonate is added, the pH will not exceed 8.3 due to the weak dissociation of bicarbonate ions, this is why I prefer it to sodium carbonate .
  4. After the bicarbonate is added to the Oxone, the solution must quickly used (<10 min). It will start to liberate free oxygen and lose strength.
  5. NaHCO3 available as baking soda. MeOH available as methyl hydrate for alcohol stove fuel. Oxone available as Oxyshock for pools (to raise active chlorine or bromine). Alkene found or derived from essential oils. Water is found distilled at a grocery store (tap water will liberate chlorine if you add Oxone, this is undesirable). DCM is available as "methylene chloride" as a paint stripper and to bond Lucite.
  6. The reflux condenser is more of a safety / odor control device than anything else, if you don't mind the smell, an Erlenmeyer flask will work fine. You can optionally run this setup with a Claisen adapter with a thermometer to monitor the temperature.
  7. If you don't have a large sep funnel (say if you're doing this 2-50x scale), find a large carboy or glass flask for shaking. Even HDPE plastic containers work well. Load the reaction mix and DCM, and shake it. Let it settle, pour off most of the water into another container, and then load the rest into your sep funnel. Get as much of the DCM as possible, then return the contents to the other container, load back into your shaker, and repeat. Yes, it's a bitch, but is it worth it for this purely OTC method? You be the judge.
  8. The amount of MeOH and H2O have been successfully reduced, look at the trials section for more information. Yields may go down, there isn't enough information yet.
  9. Good stirring is absolutely necessary, the faster the better.


Epoxidation of alkenes with potassium hydrogen persulfate
Robert Bloch, Josephine Abecassis, and Dominique Hassan
J. Org. Chem. 50, 1544-45 (1985)

Oxidation of alkenes with aqueous potassium peroxymonosulfate and no organic solvent
Weiming Zhu and Warren T. Ford
J. Org. Chem. 56, 7022-26 (1991)

The usual workup from the glycol

The more usual 65g batch is what most small-scale chemists do (after Shulgin, I suppose). In our case, we're just working with a mere 15g. Once you distilled out the DCM from the 250ml flask keep the glycol in and add:

  1. 25ml MeOH
  2. 120ml 15% H2SO4

Change the setup for reflux, and lightly reflux the mixture for 3h using a boiling hot water bath. The water bath should be kept at a light boil, and the reaction vapors will have a temp of around 65°C.

The epoxide mixture will go from a light yellowish color to more of a darker brown color.

Let it cool. Load it into a sep funnel and tap off as much of the more dense ketone as possible (both PMP2P and MDP2P are more dense than water). Now extract with 3x40ml DCM.

Wash extracts with 3x80ml 5% NaOH

For the first wash, simply rock the sep funnel back and forth. For the last two washes, shake vigorously. The sodium hydroxide washes will take away a lot of the yellowish-brown impurity color with anethole ketone, and will take away some of the reddish-brown impurity color with isosafrole ketone.

Now distill off the DCM in a water bath, or an oil bath if you're going to later vacuum or steam distill the ketone.


  1. 1kg of 15% H2SO4 is made from 150g H2SO4 and 850g H2O
  2. 1kg of 5% NaOH is made from 50g NaOH and 950g H2O
  3. Stirring is not absolutely necessary, but beneficial. The reflux provides sufficient agitation if you don't own a hotplate/stirrer.
  4. H2SO4 available as a high-strength drain opener (hardware store). NaOH is available as lye (soap shop, hardware store, grocery store). Info on finding DCM, water & MeOH is as above.

Preparation of Anethole

  1. 206.5g star anise oil fractionally distilled to 191.0g at 145-60°C at ~60mmHg
  2. Distilled again, now 178.5g at 145-48°C under ~60mmHg

Preparation of Isosafrole

  1. 250ml clear yellow sassafras oil is (optionally) fractionally distilled at 145-50°C under ~60mmHg to clear colorless safrole.
  2. Vacuum refluxed in a 2L flask (w/o stirring, no boiling stones) with 1.5wt% ground KOH flakes a la Osmium for 6 hours, the mix turns black
  3. Distilled and tossed first 15ml back with unprocessed sassafras, collected over 200ml clear colorless isosafrole at 160-62°C under ~60mmHg.
  4. Checked isosafrole by bringing to a reflux at atmospheric pressure and saw that the isosafrole boiled at 254°C. Successful conversion!


Thanks go out to Psychokitty for discovering the awesome 1991 ref which lead me to the almost as awesome 1985 ref on which this method is based. Props go out to improv chem (thank you so much for your experimentation), as well as to Osmium and Rhodium for letting me know how to figure out if I really had the goods amongst my doubts. Thanx also go out to Bright Star, Methyl Man, Ezekill, Eleusis and Shulgin for raising my awareness and inspiration.

Addendum by Chromic - regarding asarone

100 mmol asarone, 200mL MeOH, 500ml water, 24.4g NaHCO3, 72.7g Oxone. Stirred 5hrs, extracted with DCM, yield 89 mmol of crude asarone epoxide. A small amount was saved and the NMR spectra came back same as a good sample of anethole epoxide (with the exception of two additional methoxys).

15% H2SO4 rearrangement was attempted and it destroyed the molecule turning it into a crap that did form the bisulfite adduct. So, we just need to find another way to turn the epoxide to the ketone, Like LiI in ethyl acetate.

Btw, in my experience, crude anethole epoxide has been a clear yellow oil (smells nothing like licorice), asarone epoxide has been a light brown very viscous liquid (smells something like asarone, but quite different).

Addendum by Vibrating Lights - on Oxone grades

If you have the 85% [Oxone brand] then some of the other 15% is a base added by the manufacturer so the addition of the mix does not affect the pH level of the pool. You then need to adjust your amount of bicarbonate added to the rxn. Chromic uses 100% Oxone in his write up. SWIM had good yields using the 85% Oxone straight from the package. The base was not consumed until 3hrs into the rxn at which point the pH began to drop from 7.5 to 5.0. When the pH drops like that, then instead of the epoxide you are producing glycol. Tests have been run and extra base is only needed if you wish to perform thermal rearrangement. The extra base will ensure that there is no glycol presentand should be added in small amounts to maintain the pH around 6.8-7. However, if using the standard H2SO4 rearrangement the 85% Oxone mix can be used as is because the H2SO4 will dehydrate the glycol to the epoxide before it rearranges the epoxide to the ketone. Using the 85% Oxone mix and adding the amount of bicarbonate that Chromic suggest brings the pH too high and renders the epoxidation capabilities of the oxone useless. Swim knows because he had the same problems, then he saw at the Oxone technical information site that pool treatments are usually premixed with 10% Na2CO3/K2CO3. There was no color to the aqueous layers in his dreams until this was realized.

Addendum by Psychokitty

Now, lately SWIM has applied your methodology along with a few tricks of SWIM's own. Here are several suggestions (some tried, some not):

  1. Filtering the aqueous Oxone solution (prepared according to your specifications)--both diligently and patiently--in order to remove the clarifier (which is a white soot) so that one can obtain a clear colorless Oxone solution.
  2. Not adding the buffer (i.e., the sodium bicarbonate) ensures that the solution remains acidic throughout. Using methanol as the solvent, this does, according to SWIM, yield the diol instead of the epoxide, just as it does in the original biphasic reaction WITHOUT the methanol. Crude yield was somewhere above 90% and in reference to previous performic reactions, smelled and looked (somewhat reddish, viscous) like the real McCoy (according to SWIM).
  3. Several samples of the post-oxidation filtered mass revealed a considerable amount of trapped expoxide (yellowish tint). Decided to solve the problem by determining what EXACTLY the mass was soluble in. Insoluble in the following moderately dilute solutions: sodium hydroxide, sodium carbonate, sodium bisulfite. Moderately soluble in an aqueous acetic acid solution. VERY soluble (completely) in a moderately dilute HCl solution (so was the clarifier, according to further - albeit dubious - tests). Didn't try using sulfuric acid. Stupid. This will have to wait.

    I believe the methanol is what contributes to the formation of the slurry. It probably makes the salt mixture insoluble. If at the end of the reaction the methanol were distilled out, the solution might clear up a bit (if not altogether entirely) thus making the final extraction much easier (no filtration, that is).

  4. First trial with this method SWIM accidently used 1-alkene which SWIM thought had been isomerized successfully. It hadn't. After reaction was done, practically 95% of the 1-alkene was recovered unharmed. The other 5% was no doubt lost through mechanical manipulation. This reaction is VERY specific to 2-alkenes and DOESN'T seem to promote by-product formation. VERY clean.
  5. It is SWIM's opinion that the increase in temperature from 23°C to that of 36°C or above is not caused by the reaction of Oxone with the 2-alkene but through its reaction with the methanol. The resulting mild exothermic reaction thus promotes the formation of the epoxide or diol through reaction of the 2-alkene with the excess Oxone.
  6. Old paper detailing the oxidation of many substances using Oxone hints strongly that ethanol can be used in place of methanol as the solvent in this reaction. Use of isopropanol is also possible. All solvents seem to react to some degree with the Oxone in solution.
  7. According to the 1991 ref, buffering the Oxone solution before epoxidation ensures that by the end of the reaction, there is NO more oxidant left. The Oxone is therefore simultaneously decomposed by both the 2-alkene and the buffered solution. In contrast, the acidic solution remains quite stable with respect to any Oxone not used up in the 2-alkene hydroxylation.
  8. Isomerization of distilled 1-alkene with 1% KOH can be done, according to SWIM with >90% yields in under thirty minutes at 243°C.

    Experimental: Dissolve using heat and good stirring 1 gram of KOH in ethanol of methanol. Once dissolved, add 100 grams of distilled 1-alkene to mixture and increase heat to distill off the alcohol, which can be reused. Place thermometer in flask and continue to heat and stir until temperature reaches 243°C (it will hit 232°C first, linger awhile, but will soon ascend). Hold for 5 minutes, the turn off heat and let cool. Next, vacuum distill to get desired product.

    If dirty 1-alkene is used, considerable tar will form, covering up the KOH (making it inactive). Also, the stirring will be slowed immensely. Up to 500g of dirty 1-alkene was used. Stirrer stopped at one point but continued heating and refluxing ensured relatively complete conversion into the 2-alkene. For odor control, a dampened paper towel was lightly stuffed into the opening of the flask. Keep in mind that the cis-isomer of the alkene in question boils at 245°C at normal pressure. The trans boils at 250°C or so.

  9. Sodium Bisulfite could destroy any excess oxidant, but there probably isn't any left over in the methanol solvent reaction because of the methanol. Anyway, in a dilute acidic solution of dichromate, the bisulfite ion is stable and reduces the oxidant. BUUUUT, in highly concentrated acidic solution, the dichromate is still reduced but a considerable amount of SO2 is evolved from the solution. So the question is, could one reduce the excess Oxone using bisulfite at a pH of 1.3 or so?
  10. A preliminary test using the buffered epoxidation reaction, followed by hydrolysis using 15% H2SO4 yielded a mixture of products: First fraction under vac distillation was the epoxide (a light-light [lighter than the ketone] fluid oil that is almost indistiguishable from 2-alkene that does not freeze even at 0°C); second fraction, about 8°C higher than the epoxide was the ketone at the correct temp (all characteristics were on the mark); last fraction was the diol at about 10-15°C above the ketone temp (standard characteristics for those in the know). Hydrolysis of the epoxide will most certainly have to be done for longer periods of time to ensure complete conversion of the epoxide to the diol to the ketone. Crude yield was about 85% due to mechanical losses.
  11. According to the British Patent reference detailing the performic synthesis of MVK (MethylVanillylKetone) as detailed under the "Musings on the synthesis of safrole from eugenol" heading (or whatever it's called) on Rhodium's page, the diol of isoeugenol is dehydrated using a biphasic mixture of 10% H2SO4 and toluene for every 8.1 grams of starting isoeugenol used in the reaction. According to the patent, "MVK was the first and only product isolated". My guess is that any left over by-product (such as unreacted 2-alkene) is what causes the lower yields of product and the concommitant formation of some tar during the dehydration step. Possibly the use of a biphasic toluene/dilute H2SO4 solution stirred for 6 hours is the solution to this problem.
  12. Suggestion for experimental purposes a modification to the Oxone reaction as hinted in the above Brittish Patent reference wherein they detail a "through process" of acetylating isoeugenol to form acetyl isoeugenol, which is then epoxidized using a buffered (sodium acetate) peracetic acid solution, which is then neutralized with sodium bisulfite, which is then mixed with biphasic toluene/dilute H2SO4 and reacted to get in one pot MVK "as the first and only product":


  1. Add all the reactants together as per your method sans the bicarbonate.
  2. At the end of 5 hours, determine if the oxidant is spent due to the presence of methanol. If not able to due this, add the appropriate amount of sodium bisulfite in order to neutralize any excess Oxone. Keep in mind that the pH might be too acid. Determine a way to work this out. (Maybe use sodium sulfite found at the pool shop instead of bisulfite? But I think it forms a basic reaction with water, whereas bilsulfite forms an acid reaction.)
  3. Dehydrate using any one of several ways available. If the solution was buffered, and all the acid was neutralized to completion, there will be no Oxone, but plenty of Potassium sulfate, which is not very soluble in water. Trying to make a voluminous acidic solution out of this one for the purpose of dehydration is going to be difficult.
  4. Partial neutralization of the mixture, if done right, would leave an aqueous solution of potassium bisulfate. This has been shown to work well at dehydrating alcohols in aqeous mixtures. One could go with the diol reaction, neutralize the oxidant, buffer till bisulfate formation is complete (may already be significantly so), dehydrate with the methanol possibly distilling out (so as to avoid forming a methanol ether instead of the ketone - I read about this in a somewhat recent paper studying this reaction). And then either extract the solution with solvent or add water and steam distill.
  5. Or, neutralize oxidant, distill our methanol and then concentrate aqueous solution and then find a way to change the solution to 10-15% H2SO4 while of course considering the potassium bisulfate and sulfate factors. Then react as per usual using just the acid solution, or with methanol, ethanol, or toluene added.
  6. You can even go all out and start with distilled 1-alkene, isomerize as detailed above, and then add the whole sheebag directly to the Oxone mixture and epoxidize right off the bat. The KOH is in such a small amount, it probably won't make that much of a difference. If worried, however, just neutralize it with sulfuric acid or possible even better, acetic acid to form sodium acetate, which can act as a buffer.