Wednesday, January 29, 2014

TFDO Synthesis Procedure

TFDO aka methyl(trifluoromethyl)dioxirane is a very powerful reagent for the C–H oxidation of unactivated alkanes. Unfortunately, it is not commercially available and consequently needs to be prepared. Due to the volatility of TFDO, its preparation is sometimes seen as tedious if not finicky, but with the correct set-up and directions, we make TFDO without fail! It’s actually a sort of rite of passage for people working on C–H oxidation in the lab!

The preparation is slightly different from that of the original paper. (Tetrahedron, 1996, 52, 2377–2384.

Methyl(trifluoromethyl)dioxirane (TFDO): A 500-mL, three-necked, round-bottomed flask was equipped with a large stir bar and a condenser that was attached to a 25 or 50-mL receiving flask, cooled at -78 °C (acetone bath with dry ice). A hose connector must exist between the condenser and the flask to allow for pressure release/condensation (see picture 1).

The cooled (ice-water bath) three-necked flask was charged with a slurry of NaHCO3 (26.0 g) in water (26 mL), then solid Oxone® (48 g) was added to the vigorously stirred slurry of NaHCO3 (see Note below, addition takes place over 1–2 min). A lot of CO2 gas evolves during this process (see picture 2). 

After 2 min, the pre-cooled (-20 °C, freezer temperature) liquid addition funnel was placed on the three-necked flask and was quickly charged with trifluoroacetone (24.0 mL). Then, the trifluoroacetone was added within ca. 10 s. After a few seconds, the pale yellow solution of the methyl(trifluoromethyl)dioxirane (TFDO) in trifluoroacetone was collected in the cooled (-78 °C) receiving flask (see pictures 3 and 4).

After 20 min, the receiving flask was tightly closed with a plastic stopper (yellow cap). The TFDO yield was 2.0 ± 0.5% (relative to trifluoroacetone) and its concentration was determined iodometrically (0.5 mL H20, 1.5 mL glacial acetic acid, 0.25 mL saturated KI solution; addition of 0.100 mL of the TFDO solution at –78 °C; titration with a freshly standardized 0.05 N Na2S203 solution). The concentration of TFDO in trifluoroacetone ranged typically from 0.4 to 0.6 M and the volume from 4 to 7 mL. Then, the receiving flask was wrapped with aluminum foil to protect the TFDO from light and stored at –80 °C (this reagent can be stored for several months without a drop in concentration).

Important things!!

  • Wash all of your glassware with a solution of EDTA (c=0.1M) before the reaction to avoid any trace metal contaminants.
  • You need a very good stirring plate with a big magnetic stir bar
  • You can add the Oxone® in 5 g portions
  • Make sure the system is all closed before the condenser; after the condenser, there should be an open hose connector to release pressure and condense TFDO.
Finally the most important things!!

  • The quality of the Oxone® is essential (the one from Aldrich is very good) and you have to crush it into a fine powder before use.
  • The quality of the trifluoroacetone is also essential (also from Aldrich)
  • Use Millipore water

Good Luck!
Quentin & Guillaume


  1. it would be nice if someone discovered a fridge-storable small-MW dioxirane reagent that can be prepared in decent yield using a simple procedure.

    Difluorooxirane F2C=(O2) is a room-temp stable gas (and potent agent in epoxidation of perfluoroalkenes) but the preparation is atrocious. Perhaps CF3CF=(O2) could be made via CF3CO-OF, but it would involve work with F2 gas. If anybody has a better idea...

  2. Do you have any information about the thermal properties of this compound and the reaction mixture, it would seem to me to be thermally labile, hence a hazard!
    Also I would be careful grinding or milling ozone, you may set off an exothermic decomposition. I would advise obtaining DSC data.

    1. Ozone might be difficult, but DuPont sells oxone by the bucket for people to throw in their pools so there has to be DSC data somewhere.

      TFDO is volatile and a decent oxidant, and trifluoroacetone is not too good for you, so yes it will be a hazard.

    2. Damm why did I type ozone! You would have fun grinding that. Sorry.

  3. This comment has been removed by the author.

  4. ethyl and methyl esters of pyruvic acid actually form fairly stable dioxirane in situ with OXONE - it has been even proposed to use the combo as enviro-friendly low-toxicity disinfectant. But removing pyruvate esters from product would be unpleasant because they are high-boiling... But I am thinking about trying CF3COCO2Me, it costs about 2USD/gram in bulk and boils around 80C / The ethyl ester is even cheaper, about $1/gram. Vacuum transfer at 0C should do it, for the isolation of the dioxirane

  5. thank you very much for the detailed procedure. Since i am currently trying to synthesise TFDO, this post is really helpful. Some further question:
    Do you dry the TFDO-solutions over Na2SO4/MgSO4? I have the feeling that you really lose material especially during filtration, but i am not sure about the influence of water on storage and further reactions. How do you handle the material? Even with cold syringes I sometimes even can't draw up the volatile liquid and as soon as you try to estimate the right amount its gone. For larger amounts of trifluoro-acetone i use a glass syringe that stays cold at least for a few minutes long enough to transfer ~25 mL. Do you really titrate at –78°C? The Acetic Acid/Water mixture is solid at that temperature, so i'm wondering if you just cool it down for the addition of TFDO and then continue at room temperature? Have you done NMR-Titrations, e.g. Thioanisol-oxidation?
    Maybe the most puzzling question: how do you prepare these twerking dioxiranes someone mentioned in one of the last comments?

    1. By the way, 4-methoxythioanisol is considerably less smelly than thioanisol, has nicer NMR spectra, and oxidizes cleanly to the corresponding sulfone (I actually needed the sulfone as a building block, and I oxidized 4-methoxythioanisol with OXONE in MeOH)

  6. Johannes, sorry for the confusion concerning the titration. We add the TFDO solution (cooled at – 78°C) to a vial containing the potassium iodide in water/acetic acid at room temperature. We don't dry the TFDO because the reaction is not sensitive to water and I think you would loose most of your TFDO doing it. There are a few procedures in the literature to extract TFDO with DCM, but from what I heard, it's very tricky and fail most of the time. Since we keep the TFDO at – 78 °C, the water generally precipates out and stays a the bottom of the flask, so you don't dilute your TFDO that much. We generally use micropipettes to take out some TFDO and we pre-cool the tips at – 20 °C before using them. It avoids potential metal contamination, but you might loose a bit of accuracy since it's not a water solution... At the end of the day, you'll probably have to add an excess of TFDO because it is hard to add exactly what you want, especially on small scale.

  7. Milkshake, you're making a great point: a more user friendly analog of TFDO would be great and we tried to make some in the lab. Unfortunately, we haven't found a suitable replacement yet that would combine the reactivity of TFDO and a greater ease of preparation...

    1. TFDO yield is very low (2%), DMDO has even worse problem. But I have seen high yield epoxidations with dioxirane where the dioxirane was generated at 0C in biphasic mix containing methyl ethyl ketone, buffered oxone, and the olefinic substrate. I think it might be feasible to perform TFDO oxidations in situ, in biphasic conditions, using neat CF3COMe or CF3COCO2R (R=Me, Et) as a co-solvent, this would have the advantage of not having to isolate the nasty dioxirane, and it could also improve the yield

  8. This comment has been removed by the author.