Counter temporarily out of order due to NetScrape Server stupidity.
This section has been completely rewritten, as new information has been received about acid-base extraction and about extraction of DXM+guaifenesin preparations.
Please remember to always wear safety goggles when working with chemicals, and be generally careful with these procedures. My thanks to all who did research on this subject.
I'm going to present this as "kitchen chemistry" as I feel most people with adequate chemistry knowledge (and equipment) will be able to do it correctly anyway.
The older procedure for extracting DXM was to basify it with NaOH (sodium hydroxide), and filter out the precipitated DXM using a coffee filter. This tended to fail for several reasons. First, the DXM precipitate was often so fine that it went through the filter paper. Second, many syrups contain propylene glycol, and DXM free base seems to be moderately soluble in propylene glycol.
You can, of course, still use the precipitation procedure; I just don't recommend it. If you do choose to precipitate DXM, try to get actual filter paper rather than a coffee filter - it will help.
The acid-base extraction process is a common method for isolating a desired chemical from undesirable "gunk". The theory is that certain chemicals (generally, alkaloids) occur in two forms: a water-soluble complex with an acid, and an oil-soluble free base form. For example, pseudoephedrine (Sudafed), a decongestant, is usually supplied as the hydrochloride salt (pseudoephedrine HCl). It can also exist as a base, without an acid molecule (thus the term "free base"). You can convert an alkaloid from acid salt to free base (or vice versa) using a base (or acid).
The practical upshot is you take your chemical and "gunk", and lower the pH until the chemical converts to free base form and precipitates out (since it's no longer soluble in water). Now you add a nonpolar solvent (an "oily" layer) for the chemical to dissolve in, shake for a long time, and all the chemical you want is in the nonpolar layer. Discard the polar (i.e., water) layer, and you're left with a nonpolar layer full of your chemical .....
Plus anything else that might be oil-soluble. So you reverse the process, by adding an acid until the free base turns into an acid salt, and precipitates out of the nonpolar layer. Add water, shake, and you can discard your nonpolar layer.
This is the acid-base extraction, and it's very frequently used to extract the active ingredients from plants (free clue: the THC in marijuana is not an alkaloid and thus won't extract this way).
So how do we apply this to DXM? Well, it turns out that DXM is an alkaloid, and you can extract DXM from cough syrups using the same process. Furthermore, this procedure even works for DXM plus guaifenesin syrups, e.g., Robitussin DM, and generic equivalents (invariably called Tussin DM). The "DM" syrups usually only contain 10mg/5ml of DXM, so you won't get as much yield, but they're usually cheaper (and more commonly available).
Do NOT try this extraction procedure with cough syrups or formulations containing acetaminophen/paracetamol, pseudoephedrine, other decongestants, or antihistamines. (I'm working on it)
For this procedure you will need:
To speed up the process (from overnight to about 30 minutes), you will have to evaporate the solvent with heating. For this you will require:
A brief word about organic vapors. The solvents you will in all likelihood be dealing with (hexane, heptane, petroleum ether, whatever) are bad for you. Really bad for you. You do NOT want to breathe the fumes. Get it? So, if you want to speed up the process, pony up US$30.00 or so for an OSHA organic vapor gas mask (tell `em you'll be painting with oil-based paint). Sure, it's uncomfortable and looks dorky. But it sure beats brain damage! Additionally, you must do the evaporation outdoors (unless you happen to have a fume cabinet handy. And NO, the stove or bathroom fan does NOT count as a fume cabinet).
Okay, here we go:
A few comments. First, guaifenesin seems to itself convert to an oily free base layer if you add too much sodium hydroxide, so don't overdo it. Second, if you happen to have lab equipment you can of course use a separatory funnel (which is what the plastic baggie is for). Third, if you don't think you got anything, make sure the baking dish is completely dry; sometimes the DXM free base plus propylene glycol can look a lot like the glass itself.
If you want to use the precipitation method, all you have to do is add sodium hydroxide to the cough formula as described above, until the DXM precipitates out. Let it stand (or centrifuge it), and filter. The (very fine) powder that hopefully was caught by the filter paper is the DXM free base. I don't know whether the precipitation method works with DXM+guaifenesin preparations.
First I'd like to emphasize that this is still preliminary. Testing is not complete, and in particular if the method fails to remove all the acetaminophen you could be in a lot of trouble. So consider this as a starting point for further research only.
The basic principle here is differential solubility - different ingredients are soluble in different solvents. The relevant solvents here (with solubility listed if I could find it) are:
|Substance||Cold H2O||Hot H2O||Ethanol||Ether|
|DXM HBr||Soluble (<1.5%)||Soluble (25%)||Soluble (25%)||Insoluble|
|DXM free base||Insoluble||Insoluble||Soluble||Insoluble?|
|d-Pseudoephedrine free base||Slightly||Slightly||Soluble||Soluble|
|Polyethylene Glycol 400||Soluble||Soluble||Soluble?||?|
This information is from the Merck Index; I'm trying to fill in the unknowns from other sources. In particular, I'm fairly certain that DXM free base is insoluble in ether, and that d-pseudoephedrine HCl is soluble in ether.
Three steps should take care of pretty much everything: extraction with ether (to remove propylene glycol, guaifenesin, and possibly d-pseudoephedrine HCl), precipitation of acetaminophen by cooling, and precipitation of DXM by NaOH. It is possible that the removal of acetaminophen in a separate step can be avoided if NaOH doesn't cause precipitation of acetaminophen (I'm working on it).
As propylene glycol and guaifenesin are soluble in water as well as ether, the extraction with ether should probably be repeated. In particular, chilling the water would probably help by reducing the solubility of the guaifenesin.
The removal of acetaminophen is fairly straightforward; just cool the water and remove any precipitate. You want to make sure all the alcohol has been removed before this step, though. Unfortunately DXM itself becomes less soluble in cold water, so you might lose a little bit.
The final step (precipitation of free base) is the same as usual; the only caveat is that if the original preparation contained d-pseudoephedrine, some of it might have made it this far and precipitate as well. Again, I'm working on it.
I'll include procedural steps when I finish research and testing. Please remember that this is an untried procedure and may not be safe! I encourage anyone with comments to let me know what you think.
Wait a few months and this should be worked out. In particular, I think there may be a way to test for the presence of acetaminophen fairly simply. I'm not sure yet; in the mean time, please play it safe.
The DXM you extracted is in free base form, so it is theoretically possible to smoke it using a vaporization pipe. This is, however, a difficult task; if you overheat it, it starts to smell like burning plastic, and in any case it's very harsh. A few people have successfully smoked DXM freebase, but most say it's not worth it.
You can also dissolve it in alcohol, or load it into a capsule, and swallow it. In an ideal world, the hydrochloric acid in your stomach would react with the DXM to form DXM hydrochloride, which could then be absorbed. Unfortunately, we don't live in an ideal world, so it might be a good idea to form an acid salt with HCl or whatever is handy (acetic acid might work; maybe even orange juice?). The other option is to eat food with the DXM to increase both stomach acid production and lipid transport. Please note that using excess HCl may convert the DXM to dextrorphan. Incidentally, DXM itself tastes really nasty.
Or, you can use the free base DXM for further syntheses - see Section 7.5.
Not easily. One procedure starts with 3-methoxyphenethylamine and 4-methoxyphenylacetic acid (good luck finding either one at your corner hardware store). Another uses 1-p-hydroxybenzyl- 1,2,3,4,5,6,7,8-octahydroisoquinoline (assuming my translation is correct). In either case, it doesn't look particularly easy to do.
If you're still intent on trying it, I suggest you start with US or foreign patents (e.g., US patent 4613668), chem. abstracts, or journal articles.
All chemical processes in this section require pure DXM. If you do not have pure DXM, you must extract from cough formulae as above (and purify it really well). Most of these processes require significant skill, and access to lab equipment and chemicals. To my knowledge none of this is illegal (but don't take my word on it). Don't fret if your yields aren't as good as specified. Most of the procedures are from the same source (97).
This is probably the easiest by far. In fact, it's often accidental in the isolation of pure DXM. Any excess of acid (HCl or HBr) should produce dextrorphan. The primary reference for this section (97) used 48% HBr. It is possible that this occurs accidentally in some extraction procedures where HCl is used to convert DXM free base to water-soluble form. This may account for people indicating that extracted DXM is stronger than DXM in cough formulae.
These compounds would most likely have opiate activity. Unfortunately, as someone (wish I remembered who!) once put it, the isomer fairy isn't going to descend from heaven and wave her magic wand. You'd basically have to get the cross bridge to flip around (if you could do this, the hydrogens would probably conform as desired). Good luck! Personally, I don't think it can be done, at least not easily. By the time you got the lab and chemicals to do it, it'd probably be easier just to make methylfentanyl from scratch.
If you do figure out a way to do it, please don't tell anyone; nothing would bring the DEA into this faster than someone making an opiate out of DXM. You don't need to tell me either, since I don't consider opiates to be much fun. Oh, and if the isomer fairy does show up, you might as well ask her to make you some methamphetamine from Vicks Nasal Inhalers.
Several 3-substituted DXM analogs have been synthesized. A few of these actually show interesting binding and anticonvulsant activity. Table 4, on the following page, lists the analogs, their binding, and their anticonvulsant activity in rats. All data on 3-substituted analogs comes from (97). Incidentally, this article is marked as "not subject to US Copyright"; therefore I've quoted large sections from it. Curiously enough, the research was sponsored by NIDA (the National Institute on Drug Abuse). ED50 Rats refers to the effective dose for anticonvulsant activity; % Rats refers to the percentage of rats protected. Sample size was 10 rats.
Some comments on the table. Both dextromethorphan and the N(CH3)2 derivative lost anticonvulsant activity at higher doses. The NH2, OEt, and O2-Pr derivatives all showed no indication of psychotomimetic activity at anticonvulsant doses. Most showed little ability to displace [3H]TCP. Generally speaking, I think it's safe to say that the [3H]TCP binding site is the NMDA open channel PCP1 site, and that the [3H]DXM binding is occurring to DXM's high affinity sites (sigma1 and PCP2). The authors do not address the PCP2 site.
My guess is that the discrepancies between [3H]DXM binding affinity and anticonvulsant activity relate to different binding at sigma1 and PCP2, and that the anticonvulsant activity comes from the sigma1 activity. As far as any recreational use of these derivatives goes, I have no idea. Potentially, the NH2 derivative might show effects limited to sigma1 activity, and the OEt and O2-Pr derivatives might show sigma1 and PCP2 activity. It doubt any of the above are specific for PCP2; the closest would be the H "derivative". This is all scientific wild-assed guessing; there's not much data on PCP2 (or the sigma receptors for that matter).
|3-Position Substitution||IC50 [3H]DXM||IC50 [3H]TCP||ED50 Rats||% Rats|
|OCH3 (DXM)||0.59M (0.12)||2.0M (0.6)||38mg/kg||70|
|OH (dextrorphan)||7.7M (0.9)||1.2M (0.7)||5mg/kg||90|
|NH2||45% at 10M||7.8M (1.4)||25mg/kg||100|
|NHCH3||3.6M (1.4)||43% at 10M||0|
|N(CH3)2||4.4M (0.9)||45% at 10M||40mg/kg||40|
|Cl||1.1M (0.4)||5.5M (1.5)||10|
|NCS||1.5M (0.3)||60% at 10M||0|
|H (i.e., nothing)||1.3M (0.3)||53% at 10M||0|
|O-Et (ethyl)||0.42M (0.06)||75% at 10M||5.6mg/kg||90|
|O-2-Pr (2-propyl)||0.88M (0.18)||59% at 10M||3.9mg/kg||90|
|O-n-Bu (n-butyl)||1.5M (0.4)||58% at 10M||40|
|O-Bz (benzyl)||3.1M (0.6)||39% at 10M||30|
So if you want to go about synthesizing any of these, I don't believe it would be illegal (I could be wrong). I wouldn't advise taking any of them, of course; in particular, there's been no LD50 determination. The authors doubt that the NCS derivative even gets to the brain. If it did get to the brain, it would likely bind irreversibly. You don't want that (imagine tripping for three months).
I haven't had time to check these extensively for typos, so you might want to check out the original article. Note that the authors made some typos of their own (nothing in the chemistry though). If you're reading this as a hypertext document, I put in plenty of cross-reference links; if you get confused about a referred intermediate, you can jump to it.
Dextrorphan was obtained in quantitative yield by a general method of O-demethylation of dextromethorphan HBr [obtained from Sigma/Aldrich] with 48% HBr, and was found to be identical by TLC, mp, and 1H NMR to an authentic sample. By using a modification of the procedure by Conrow and Bernstein [Steroids 1968, 11, 151-164], [dextrorphan] (16.10g, 62.3 mmol) was dissolved in hot acetone (500mL); Am-ex-OL (15.51g, 64.4mmol, Aldrich Chemical Co.) and K2CO3 (17.13g, 123.9mmol) were added. The reaction mixture was allowed to stir at reflux, under an atmosphere of argon, overnight. After cooling, the reaction mixture was extracted from H2O (500mL) with benzene (1x500 and 2x250mL) and removing the solvent in vacuo, to give 27.51g (96%) of [this intermediate] as a white glass which was homogeneous by TLC. Recrystallization in EtOAc gave pure [this intermediate] as white crystals: mp 138-139 C; EIMS m/z 461; [alpha]D +59.5 (c 1.03, acetone). Anal. (C31H31N3O) C,H,N.
[The above intermediate] (10.0g, 21.7mmol) was placed in mineral oil (100mL, light white oil, Sigma Chemical Co.) under a stream of argon, and the vigorously stirring reaction mixture was carefully heated in a sand bath to 330-350 C. Complete conversion of [the above intermediate] to one major product spot occurred in 8-10 h. (Note: This reaction proceeds significantly slower at lower temperatures and will rapidly decompose at temperatures that exceed 360 C). The yellow reaction mixture was allowed to cool to room temperature and suction filtered through a pad of silica gel. The mixture was eluted with 700mL of ether (to remove all mineral oil). Then, the receiving flask was changed, and the product was eluted with CHCl3/MeOH/NH4OH (800mL, 90:10:1) to give 5.95g of (60%) yellow foamy product that was homogeneous by TLC. This intermediate was not crystalline. It was characterized spectrally and taken to the next step without further purification: IR 1690 (C=O) cm-1; EIMS m/z 461.
Sodium hydroxide (4N, 30mL) was added to a solution of [the above, second intermediate] (4.89g, 10.7mmol) in absolute EtOH (100mL), and the reaction mixture was allowed to stir at reflux, under an atmosphere of argon. After 18h, TLC showed a complete loss of [the second intermediate]. The reaction mixture was cooled in an ice bath and carefully acidified to pH 2 with concentrated HCl. Additional 1N HCl (100mL) was added and the reaction mixture was allowed to stir at reflux, under an atmosphere of argon for 1.5h. After cooling in an ice bath, the aqueous reaction mixture was extracted with ether (1x200 and 2x100mL). The ether fraction was washed with 1N HCl (1x200mL) and the combined aqueous fraction was neutralized to pH 9 with NH4OH. Extraction with CHCl3 (1x200mL and 1x100mL) followed by CHCl3/MeOH (4:1, 1x100mL) and removal of solvent in vacuo resulted in 2.53g (93%) crude [NH3 derivative] as a white foamy free base. The free base was dissolved in a minimal volume of hot MeOH and acidified with a saturated solution of HCl in 2-PrOH. Addition of ether resulted in an oily product, but removal of solvent followed by recrystallization in MeOH/ether resulted in 1.0g (34%) of [NH3 derivative] as white crystals, 280°C dec. The mother liquor was neutralized and purified by flash column chromatography (CHCl3/MeOH/NH4OH, 95:5:1) to give 0.84g (31%) pure base as a white foam: IR (free base, CHCl3) 3400 cm-1 (two sharp bands, NH2); 1H NMR (CDCl3) delta 6.90 (d, J=8.4Hz, 1H), 6.60 (d, J=2.4Hz, 1H), 6.50 (dd, J=8.4,2.4Hz, 1H), 3.55 (brs, 2H), 2.60 (s, 3H); CIMS m/z 257 (M+1); [alpha]D +21.7 (c 0.53; MeOH). Anal. (C17H24N2*2HCl) C,H,N.
Formaldehyde (37%, 0.30 mL) and succinimide (0.45g, 4.5mmol) were added to a solution of [NH3 derivative] (0.76g, 3mmol, free base) in absolute EtOH (8mL). The reaction mixture was allowed to stir at reflux, under an atmosphere of argon, for 2h. The volatiles were evaporated, the residue was redissolved in DMSO (3mL), and NaBH4 (0.18g, 4.7mmol) was added, at room temperature. The reaction mixture was warmed to 100 C, allowed to stir for 15 min, and cooled. Quenching with H2O (25mL) and extraction with CH2Cl2 (3x25mL) was followed by washing the combined organics with H2O (3x20mL), drying (Na2SO4), and evaporating to give 0.50g (63%) of an off white foam that was nearly homogeneous by TLC. The foamy free base was dissolved in a minimal volume of hot MeOH, and oxalic acid (0.33g, 3.7mmol, 2 equiv) was added (pH 4). The crystalline product was isolated and recrystallized in MeOH to give 0.51g (61%) of [NHCH3 derivative]: mp 187-193 C; IR (salt, KBr) 2400 cm-1 (broad NH+); 1H NMR (CDCl3) delta 6.92 (d, J=8.4Hz, 1H), 6.50 (d, J=1Hz, 2.4H), 6.43 (dd, J=8.4,2.4Hz, 1H), 6.50 (d, J=1Hz, 2.4H), 6.43 (dd, J=8.4,2.4Hz, 1H), 2.78 (s, 3H), 2.35 (s, 3H); EIMS m/z 270; [alpha]D +22.0 (c 1.08, MeOH). Anal. (C18H26N2*C4H4O8*1/4H2O) C,H,N.
Formaldehyde (37%, 2.4mL) and NaBH3CN (0.60g, 13mmol) were added at 0°C to a solution of [NH3 derivative] (0.52g, 2mmol, free base) in MeCN (10mL). The reaction mixture (pH 12) was allowed to stir at 0°C for 5 min and then allowed to warm to room temperature. After a reaction time of 30 min, glacial HOAc (0.3mL) was added and the reaction mixture (pH 6) was allowed to stir at room temperature for another 45 min. The volatiles were removed in vacuo, and the residue was extracted with 2N KOH (1x20mL) and ether (4x10mL). The combined organic fraction was washed with H2O (1x10mL), dried (Na2SO4), and evaporated to 0.53g of a pale yellow oil (93%) which was nearly homogeneous by TLC. The crude free base was dissolved in hot 2-PrOH and added to a solution of D-(l)-tartaric acid (0.57g, 4mmol) in hot 2-PrOH. The crystalline [N(CH3)2 derivative] was carefully isolated under an atmosphere of argon (hygroscopic): mp 96-99 C; 1H NMR (CDCl3) delta 3.15 (s, 6H); EIMS m/z 284; [alpha]D +5.73 (c 0.96, MeOH). Anal. (C19H28N2*C6H9O9*1/4 H2O) C,H,N.
Acetonitrile (16.0mL), tert-butyl nitrite (0.8mL, 90%, 5.4mmol), and dried CuCl2 (0.68g, 5.0mmol) were combined and warmed to 60 C, in an argon-purged round-bottom flask. A solution of [NH3 derivative] (1.0g, 4mmol, free base) in MeCN (8.0mL) was added dropwise, via addition funnel, over 10 min, and the reaction mixture was allowed to stir at this temperature for 2h. The reaction mixture was cooled and poured into a separatory funnel into which 10% Na2CO3 (50mL, w/v) was added. The aqueous layer was removed and the organic layer was saved. The aqueous layer was then washed with EtOAc (3x30mL), and the combined organic fraction was washed with H2O (1x25mL), dried (Na2SO4), and evaporated to 0.98g (89%) of crude [Cl derivative] as a dark oil. Purification by gradient flash column chromatography (CHCl3/MeOH/NH4OH, 95:5:1/90:10:1) yielded 0.59g (55%) of pure [Cl derivative] as the free base. The free base (0.48g, 1.7mmol) was dissolved in a minimal volume of 2-PrOH and was added to a solution of 0.21g fumaric acid (1.8mmol) in 2-PrOH. Addition of anhydrous ether resulted in crystalline [Cl derivative]. Recrystallization in 2-PrOH gave 0.41g (56%) of pure [Cl derivative], mp 136-141C. (Note: if crystallization was difficult and the crude salt was dark, boiling in 2-PrOH with charcoal followed by filtration over Celite facilitated isolation of pure product.) IR (KBr, salt) 650 cm-1 (C-Cl); 1H NMR (CDCL3) delta 7.25 (d, J=1.9Hz, 1H), 7.14 (dd, J=6.3,1.9Hz, 1H), 7.08 (d, J=6.3Hz, 1H), 2.62 (s, 3H); EIMS m/z 275,277 (M+2); [alpha]D +19.6 (c 0.53, MeOH). Anal. (C17H22NCl*C4H4O4*2H2O) C,H,N.
A solution of [NH3 derivative] (0.52g, 2.0mmol, free base) in pentene-stabilized CHCl3 (45mL) was added to a solution of NaHCO3 (0.64g, 2.8mmol) in H2O (20mL) at 0 C. The biphasic reaction mixture was allowed to stir under an atmosphere of argon, at 0 C for 10 min. Freshly distilled thiophosgene (200 microL, 2.2mmol) was added and the reaction mixture was allowed to stir at 0 C for 10 min and room temperature for 30 min. The organic layer was removed, and the aqueous layer was extracted with CHCl3 (2x20mL). The combined organic fraction was washed with H2O (1x20mL), dried (Na2SO4), and evaporated to 0.60g (98%) to an orange foam. The crude free base was dissolved in a minimal volume of 2-PrOH and acidified with a saturated solution of HCl in 2-PrOH (pH 4). Careful addition of anhydrous ether resulted in 0.59g (88%) of crystalline [NCS derivative], mp 260°C dec; IR (CHCl3) 2160 cm-1 (br, NCS); CIMS m/z 299 (M+1); HRMS (M+) calcd for C18H22N2S 298.1504, found 298.1476; [alpha]D +19.1 (c 0.68, MeOH). Anal. (C18H22N2S*HCl*3/4 H2O) C,H,N.
A modification of the procedure described by Reden et al. [J. Med. Chem 1979, 22, 256-259] was used beginning with the addition of 5-chloro-1-phenyl-1H-tetrazole (1.08g, 6.0mmol) and anhydrous K2CO3 (1.35g, 9.8mmol) to a solution of [DXM - note, not dextrorphan] (1.29g, 5.0mmol) in dry DMF (25mL). The reaction mixture was allowed to stir at room temperature, under an atmosphere of argon, for 18 h. Extraction from H2O (20mL) with ether (3x20mL) was followed by washing the combined ether fractions with 15% NaOH (1x15mL) and then extracting the organic phase with 1N HCl (3x20mL). The ether layer was discarded, and the aqueous layer was neutralized with concentrated NH4OH to pH 9, extracted with ether (3x20mL), dried (Na2SO4), and evaporated to 1.65g of a white glass (100%) that was homogeneous by TLC and was taken to the next step without further purification: IR (CHCl3) 1660 (C=N), 1600 (Ar) cm-1; 1H NMR delta 7.50 (d, J=8Hz, 1H), 7.47 (m, 2H), 7.14 (m, 6H), 2.37 (s, 3H); CIMS m/z 402 (M+1).
A modification of the procedure described by Reden et al was used to convert [above intermediate] to [H derivative]. A mixture of [above intermediate] (1.55g, 4.7mmol) in glacial HoAC (40mL) and 10% Pd/C (1.55g) was hydrogenated (40psig, 40 C) for 96 h, when the reaction was determined to be complete by TLC (THF/hexane/NH4OH 1:1:0.1). The reaction mixture was filtered over Celite and evaporated to 1.16g (100% crude) of a clear oil. Purification by flash column chromatography (THF/hexanes/NH4OH, 1:1:0.1 to THF/NH4OH, 9:1) afforded 0.60g (53%) pure [H derivative], as the free base. The oily free base (0.30g, 1.24mmol) was dissolved in MeOH and added to a solution of fumaric acid (0.15g, 1.24mmol) in MeOU. The volatiles were removed in vacuo, and the salt was recrystallized in 2-PrOH/ether to give 0.30g (68%) of [H derivative]: HRMS (M+) calcd for C17H23N 241.1830, found 241.1825; [alpha]D +14.9 (c 1.15, MeOH).
[DXM - note not dextrorphan] was converted to the free base by extraction from aqueous NH4OH (20% v/v, 25mL) into CHCl3, followed by drying in vacuo. The crystalline free base was dissolved in freshly distilled (from P2O5) dichloroethane (65mL); K2CO3 (11.0g, 80mmol) and ACE-Cl (11.4g, 80mmol) were added at 0 C, under an atmosphere of argon. The reaction mixture was warmed and allowed to stir at reflux for 6h. Cooling was followed by filtration and removal of solvent in vacuo. The residue was dissolved in MeOH (60mL) and allowed to stir at reflux for 1h. Evaporation and recrystallization from EtOAc gave 5.29g (90%) of pure [intermediate]: mp 250°C (lit. mp 249-250.5°C [J. Pharm. Sci. 1980, 69, 1447-1448]).
[The above intermediate] (15.3g, 52.2mmol) was converted to the free base by extraction from aqueous NH4OH (20% v/v, 100mL) into CHCl3, followed by drying in vacuo. The free base was dissolved in ethyl formate (200mL) and formic acid (100%, 150 microL) was added. The reaction mixture was allowed to stir at reflux overnight. Evaporation gave the crude N-formyl-3-methoxymorphinan as a clear gum; TLC showed one major product spot, CIMS m/z 286 (M+1). The intermediate gum was dissolved in CH2Cl2 (100mL), and a solution of BBr3 (20mL, 209mmol) in CH2Cl2 was added via addition funnel, over 10 min, at 0°C under an atmosphere of argon. Five minutes after the addition of BBr3, the reaction was complete by TLC. The reaction mixture was poured onto a slurry of NH4OH (100mL, 400g of ice) and stirred for 20 min. The reaction mixture was then poured into a separatory funnel, the organic phase was removed, and the aqueous phase was extracted with CHCl3/MeOH (4:1, 3x100mL). The combined organic phase was washed with H2O (1x250mL), dried (Na2SO4), and evaporated to 12.58g (89%) of [this intermediate] as a white foam: IR (CHCl3) 3260 (OH), 1650 (NCHO) cm-1; 1H NMR (showed rotamers) delta 8.15 (s, 1H); 7.99 (s, 1H); 6.94 (t, J=6.2Hz, 2H); 6.81 (s, 1H); 6.80 (s, 1H); 6.69 (dd, J=4.2,8.4Hz, 2H); 6.26 (brs, 1H); 6.24 (brs, 1H); CIMS m/z 272 (M+1).
A reaction mixture of [above intermediate] (3.0g, 11mmol), K2CO3 (7.7g, 110mmol), and 2-bromopropane (7.0mL, 75mmol) in dry DMF (15mL) was allowed to stir at 60°C overnight. The reaction mixture was cooled, filtered, diluted with H2O (25mL), and extracted with ether (3x25mL). The ether layer was washed with H2O (3x10mL) and evaporated to a clear oil which was homogeneous by TLC and taken to the next step without further purification, CIMS m/z 314 (M+1). A slurry of LiAlH4 (1.60g, 40mmol) in dry THF (20mL) was prepared in an argon-purged three-necked round-bottom flask, at 0 C. The oily intermediate described above [in this paragraph] was dissolved in THF (10mL) and added dropwise to the reaction mixture, under an atmosphere of argon. The ice bath was removed and the reaction was complete in 30 min. The reaction mixture was cooled to 0°C and carefully quenched with H2O (1.6mL), aqueous NaOH (1.6mL, 15% w/v), and H2O (4.8mL). The lithium salts were filtered and washed with ether. Evaporation of the filtrate and drying in vacuo gave 25.9g (87%) of the free base of [2-Pr derivative] as a clear oil. The free base was dissolved in MeOH and acidified with a solution of fumaric acid (1.0g, 10mmol) in MeOU, addition of ether resulted in 3.29g (91%) of crystalline [2-Pr derivative]: mp 181-185 C; 1H NMR (CDCl3) delta 6.83 (m, 3H), 4.56 (m, 1H), 2.90 (s, 3H), 1.28 (d, J=3.1Hz, 6H); CIMS m/z 300 (M+1); [alpha]D +22.75 (c 1.09, MeOH). Anal. (C20H29NO*C4H4O4) C,H,N.