Oxycodone and Oxymorphone from Codeine

Oxycodone/morphone from codeine the hard way by (Samson)

Here is a really roundabout way of getting from codeine to oxycodone/oxymorphone. It starts with the method laid out by Rappaport and Barber in J. Med Chem, 1975, 18(11) 1074-1077 for getting thebaine from codeine:

This method goes codeine -> codeine 6-methyl ether (CME) --> thebaine.

CME: An excess of potassium hydride (300 mol) was washed with hexane (3 X 50 ml) and then suspennded in THF (50 ml). With stirrng under a nitrogen atmosphere, a solution of codeine (1.00 g, 2.33 mmol) in 50 ml THF was added to the KH suspension over a period of 2hr and stirred for an additional hour. Methyl iodide (0.43 ml, 6.66 mmol) was added to the mixture rapidly and the reaction was quenched after 90 sec with 20 ml of 1 N NaOC2H5 in ethanol. Water (50 ml) was added and the solution evaporated to remove organic solvents. The resulting aqueous mixture was extracted with chloroform (4 X 50 ml), and the chloroform extracts were washed witha small portion of water, dried over MgSO4 and evaporated to give a crude solid which was crystallized from EtOH to give 0.87g CME (Yield 83%).

Thebaine by oxidation with gamma-MnO2: A solution of CME (313 mg, 1.00 mmol) in THF (10ml) was shaken vigorously with gamma-MnO2 (440 mg, 5.0 mmol) under a nitrogen atmosphere at room temperature. Further portions of gamma MnO2 (440 mg, 5.0 mmol) were added at intervals of 1, 3, 5, and 10 hr. After 24 hr, the black mixture was filtered through a fine sintered glass funnel, the residue was washed with THF (4 X 50 ml) and the washings were were combined with the original filtrate to and evaporated to give 207 mg of crude thebaine. The MnO2 was then washed with methanol (4 X 30 ml) to yield an additonal 119 mg of crude thebaine. The two fractions were combined and crystallized from ethanol to give thebaine (251 mg, 80% yield).

Now, from the _Pharmaceutical Manufacturing Encyclopedia_ (further info on this volume by request), a procedure is given for converting thebaine to oxymorphone with oxycodone as an intermediate. It is a synopsis of US Pat 2,806,033 (1957):

Thebaine is dissolved in aqueous formic acid and treated with 30% H2O2; neutralization with aqueous ammonia gives 14-hydroxycodeinone. It is hydrogenated to give oxycodone. 90 ml of concentrated HBr are heated to 90°C. 9 grams of 14-hydroxydihihydrocodeinone (oxycodone) are then added under stirring and the mixture is is quickly heated to 116°C and kept at this temperature under reflux for 20 min., with continued stirring. The resulting brown solution is diluted with about 90 ml of water and chilled with ice. Aqueous 10% NaOH is now added to alkaline reaction and the liquid is extracted 3 times with 100 ml portions of chloroform. The layers are separated and the aqueous phase is filtered and acidified with conc. HCl, treated with charcoal and filtered.

The filtrate is treated with conc aqueous ammonia until the mixture gives a pink color on phenolpthalein paper. The liquid is extracted 7 times with 100 ml chlorform portions, the extracts are combined, dried with anhydrous Na2SO4 and evaporated. The residue is dissolved in ethanol by refluxing and the ethanol evaporated nearly to dryness. 100 ml benzene are then added, the mixture refluxed for 1/2 hour and set aside for crystallization. After cooling [oxymorphone] is collected by filtration. 2.3 g of a white crystalline powder are obtained...

Notes: The _Encyclopedia_ is notoriously circumspect about the synths it gives. For instance, the procedure given for dextroamphetamine starts with dl-amphetamine and describes how to resolve the isomers. Thanks! I don't know if that makes the information therein "unreliable", but there are surely many important blanks left to be filled in (which I don't know enough to do). The Merck Index says of oxycodone, "Prepn by catalytic reduction of hydroxycodeinone, its oxime, or its bromination products or by reduction of hydroxycodeinone with sodium hydrosulfate". This seems to be in step with the the information given in the _Encyclopedia_. Also from the Merck, oxymorphone is obtained from oxycodone by "boiling with concd. aq. hydrobromic acid", the patent listed above being cited. (also see Weiss, J. Am Chem Soc 77, 5981 (1955)).

I imagine there are ways of getting from codeine to 14-hydroxy-whatever without going through thebaine. Thebaine is the best starting material given its reactivity with hydrogen peroxide. On the other hand, it is strictly controlled, has no therapeutic use (you won't find it in any dead grandmother's medicine cabinet...), and it one of the more scarce opium alkaloids. Whether or not pharmaceutical manufacturers use thebaine as a starting material on a large scale I have not been able to verify. Older literature is rife with whining about the scarcity of thebaine and the need for new ways to get from more abundant alkaloids to 14-hydroxy compounds and oripavine. But over the last 20 years or so, Papaver braectatum, a high-thebaine poppy has increasingly been cultivated in Turkey, and the French pharmaceutical giant Sanofi has genetically engineered a very high thebaine variant of somniferum (up to 20% thebaine). On the latter point, it is interesting to note that France now surpasses Turkey in (licit) opium production (only topped by India and Australia). Does this mean that thebaine is flooding the labs? I don't know.

Posted by Lone Ranger

After looking over Rhodium's site for the 100th time, and sending him numerous complaints about small details on some of the posted methods, I would like to publicly acknowlege that his is far and away the best site of it's kind, and the best maintained.

And so I offer this much requested synthesis:

14-HYDROXYCODEINONE: The same product obtained from the 30% hydrogen peroxide treatment of thebaine in acetic acid can also be obtained directly from codeine.

30 grams of codeine in 80 ml water with 25 grams acetic acid, is treated cold with a solution of 20 grams sodium dichromate in 25ml water. An oily precipitate forms which becomes crystal after short warming. The mixture is heated to 80°C with stirring until all the solid goes into solution; the temperature rises spontateously to 90°C. After short standing, the base is precipitated from the cold solution as the dichromate by adding an excess of chromic acid solution. Crystallize from alcohol with a little chloroform as plates, decomposes at 275°C. Stable to strong acid or alkali. Soluble in chloroform, ethyl acetate, ligroin. Sparingly in alcohol, insoluble in ether or water.

Simple hydrogenation over Pt, Pd, or Ni would satutate the recovered and purified compound to the potent dihydro version. Hydrogenation over Pt or Pd in dilute acetic acid give 93% to 97% yields. This product would be 14-hydroxydihydrocodeinone or Oxycodone, either as the dichromate or possibly acetyl salt.

What I like about this procedure is that it is simple and fast and uses only very available reagents.

What I don't like about it is that no yields are stated.


  1. Chemistry of the Opium Alkaloids, L. small and R. Lutz, US Gov printing office, 1932, pg 259
  2. Merk and Co, DRP 411530
  3. Frdl 12,750
  4. Houben 4, 572

Q: What is "Simple hydrogenation"?

A: The substrate is dissolved in a protonic solvent, such as CH3COOH, and placed in a reaction vessel with a catalytic amount (small) of Pd or Pt on a suitable substrate (ie 5%Pd/C), and shaken or stirred magnetically. H2 uptake is monitored (A graduated cylinder inverted in a water bath). The reaction is complete when H2 uptake ceases. The hydrogen is measured by reading the water level in the inverted cylinder. No heat is needed, as this is a simple ring saturation. Raney nickel is OK, but I do not have (off hand) the yield for this catalyst. It is 97% with Pd, 93% with Pt. Another nice reaction, as nothing happens until the stirrer is turned on. At that time the water level rises (to replace the comsumed H2) at an almost visible rate.

Codeine --> Oxycodone

by dope_amine

Codeine -> 14-Hydroxycodeinone

Dissolve your codeine base in the acid solution (acetic, we're following the method given in Lone Ranger's procedure above, so use those amounts) in a rb flask and set it in an ice water bath with stirring. Now that it's nice and cool, slowly pipette in the dichromate solution. The solution immediately turns a milky orange/yellow. Now remove the ice and set it in your oil bath on a hotplate and throw your water cooled condenser on top. I used a temperature controlled hotplate so I just dialed the temp in to ~83°C and made sure it stayed there for ~20 min. If you stick a thermometer in your oil bath you should be able to do about the same. Heat is bad for opiates, but I have not found that letting it cook at 80-90°C for 30 min to be terribly detrimental. Anywayz, you can tell that the reaction is complete when there is no more color change. The new color should be a dark brown/red. Pipette pour your solution into a sep. funnel, wash your flask with a bit of water and pour that in too. There will be some black insoluble crap left in the flask. Don't worry about that shit. Now basify with ammonia. You don't have to be exact, just make sure the pH is taken to above 10 and then extract three times with chloroform. If you get an emulsion, just deal with it by adding a butload of chloroform and only taking the unemulsified chloroform with each separation. Take your chloroform solutions and back extract with water twice to make sure you didn't bring nasty Cr with your product. The product is yellow when disolved in a large amount of solvent and dark brown/red when concentrated. You just converted codeine's hydroxy group to a ketone and put a hydroxy under the nitrogen bridge. Evaporate off your solvent.

14-Hydroxycodeinone -> Oxycodone

Next is to reduce that double bond which is VERY EASY if you have a source of hydrogen. As far as other equipment, you need two syringes with the plunger removed and needles attached. To each you attach a balloon the the plunger end. Then, you need some sort of thick rubber to cover the mouth of your flask, but can be penetrated (air tight) by your needles. Ok, so you throw in like 3 molar % (of initial moles of codeine) of Pd/C (either 10% or 5% Pd on Carbon) into your flask which contains your goodies in a fairly concentrated solution of ~8% acetic acid in MeOH (Not super concentrated, but definately not super dilute) with STIRRING. Then you fill your ballons through the needles, purge your flask air space with hydrogen by inserting both needles through the rubber diaphram and insert a third needle to allow an out flow. After it has purged for a while pull out the purge needle and make sure that the hole it created is now sealed. Refill your balloons and reinsert them. Let it stir overnight. I have done this on a few gram scale so there wasn't a lot of hydrogen absorbed. If you were doing a tens of grams scale, you would need to refill the balloons several times during the reaction or you may still need to do this if you have bad leaks. Ok, so filter your solution. It's about the same color, maybe a little bit lighter, but don't worry cuz it worked. Again, basify, shake with chloroform and extract 3x. Evaporate, disolve in acetone and gas or drip in HCl to crystalize. Your crystals may turn to goo when exposed to air during filtration if you use wet HCl. Don't worry about it, just realize you are gonna have to drip an aqueous oxy solution into your nose instead of getting to snort bumps to get high. Do not over acidify during crystalization! When your solution turns from brown to clear, stop adding acid and filter. If you overacidify, your crystals will turn to brown goo balls in your flask. They won't be destroyed but if you dried and snorted 'em the leftover HCl's might burn your nose a bit.

This procedure is very easy and the whole shebang except for the work-up could be done in a day no sweat. The yields of both steps are very high yielding. I have taken an H-NMR of the compounds at each step of the way and found them to be highly pure with no observable starting material in their spectrums. So get to work! Next time I'll go over the conversion of oxycodone to oxymorphone which is super potent and this is even more simple of a reaction. None of the reagents in either synthesis are suspicious so fuckin' chill.