• N&PD Moderators: Skorpio | someguyontheinternet

My opioid chemistry blog

Opiophiliclab

Greenlighter
Joined
Apr 21, 2025
Messages
6
Hello dear BL members. I have recently started writing a blog that is all about the chemistry of opioids. From SAR discussions, to synthesis routes, to my own derivatives designs, you'll find various interesting things written by someone who is both a chemist and a connoisseur of opioids. I have taken care to write down the information in as scientific a way as possible (this is especially true for the blog article "The Chemistry of Opioids") and also in very detailed ways. Something that I always resented very much about synthesis descriptions that one can read in patent files is the oftentimes disorganized manner in which the route is explained (on top of the total lack of step-by-step pictures) and the assumption of the authors that the readers always understand the "why" behind the authors' actions. This is why I will always explain why I have done something the way I have done it in articles where I describe synthesis routes (and if I forgot it at some point, then please write it in the comments).

I hope you'll find my blog interesting and informative. Please keep in mind that irrespective of how much I might know, I consider myself a lifelong student first and foremost, so I'm always open on being corrected and most importantly to learn from YOU (yes, that includes the newbie chemist).

 
Last edited:
What I don't understand is if you demethylated DXM you would end up with DXO not levorphanol.
 
What I don't understand is if you demethylated DXM you would end up with DXO not levorphanol.
No because DXM is already levorotatory. DXO is something you end up with after treating racemorphan with a dextrorotatory reactant.

EDIT:
I just noticed that I forgot to incorporate the stereoisomerism in my skeletal formulas. I will update the wedged and dashed lines at a later time. Until then, please keep in mind that the ethyl bridge, C9 and H14 are supposed to have wedged bonds to indicate orientation towards the viewer.

EDIT #2:
Ignore the first line. I confused the compound with something else.
 
Last edited:
I knew someone who would O-demethylate DXM and can assure you that DXO was the product. It's even in the name i.e. dextromethorphan wherase the opioid is levorphanol.

I actually chased down the modern synthesis of DXM because I wondered if they ended up with 50% of each enantiomer. It turned out not. For the step from which the isomerism is born is carried out using a chiral catalyst. Clever, because otherwise the whole production facility would have ended up looking like a prison. I say that as someone who has been to the MacFarlan Smith facility in Edinburgh where medical diamorphine is produced and it's every bit as secure as a prison - which must surely be a huge extra cost.

I would be taking a closer look at synthetics that fall into the cracks as far as legal status is concerned. There are a few examples that are partial agonists or which don't have a favourable MOR:KOR relative affinity (making them unpleasent thus not subject to abuse) and asking how they might be modified.
 
I knew someone who would O-demethylate DXM and can assure you that DXO was the product. It's even in the name i.e. dextromethorphan wherase the opioid is levorphanol.
Pardon, I confused the compound with something else. How unprofessional of me. As far as the product of my synthesis is concerned the product is most definitely levorphanol and not dextrorphan because a) my synthesis is actually a modified version of the levorphanol synthesis that can be found in this patent file here (chromatographic purity of the author's levorphanol was 99.75%), b) I consume the drugs I synthesize and the effect was not at all dissociative. In fact, with the dose I took I must have been tripping hard but that was not the case (also, the DOA was totally different), c) the solubility profile of DXO is different from the profile of Levorphanol and finally d) Levorphanol is manufactured by the pharmaceutical industry through conversion of DXM into Levorphanol due to economic reasons (even though I never produced it, my colleague who worked for a competitor synthesized it out of DXM). That is something I invariably noticed too when I tried to synthesize Levorphanol by using Phenanthrene as the starting material as I noted in my blog article. The yield was dirt poor.

Also, dextrorphan is industrially made by enantioselectively reacting racemorphan with a dextrorotatory reactant and not by SEARing DXM. It's standard procedure.
 
Last edited:
Phenomorphan is a highly potent drug due to the N-phenethyl group, which boosts affinity to the μ-opioid receptor, and so phenomorphan is around 10x more potent than levorphanol, which is itself 6-8x the potency of morphine. Other analogues where the N-(2-phenylethyl) group has been replaced by other aromatic rings. are even more potent, with the N-(2-(2-furyl)ethyl) and the N-(2-(2-thienyl)ethyl) analogues being 60x and 45x stronger than levorphanol, respectively.
 

In the 1950s someone tried various different N-substitutions of levorphanol. Some a lot more potent than their parent. However, it's important to realize that they only used animal models and in related compounds we have seen modification of the N-substituent increase every metric available from Ki to EC50 to LogP and whatever else you can think of that would STRONGLY suggest that the modification would result in a much more potent compound... only to discover that in man, it was no more potent. I speak of oxymorphone and N-phenylethylnoroxymorphone in this case. So while uncertain, those digitized images in the link are sort of interesting.

But you still have to obtain your norlevorphanol from somewhere.

While people have certainly improved on the original Grewe et al. synthesis, I would suggest that the KEY improvement is that chiral catalyst I mentioned. If it's possible to produce enantiopure dextromethorphan, it's equally possible to use a chiral catalyst to produce levomethorphan.


As you can see from the above, the intermediate is resolved but I'm sure I've read earlier works that used small molecule chiral catalysts.

But even so, unless those potent derivatives I've linked to are as potent in man, it isn't a great target. I mean, you can quite legally obtain the immediate precursor to things more potent than levorphanol. Then I guess you could also play with the N-substituants.
 
Last edited:

In the 1950s someone tried various different N-substitutions of levorphanol. Some a lot more potent than their parent. However, it's important to realize that they only used animal models and in related compounds we have seen modification of the N-substituent increase every metric available from Ki to EC50 to LogP and whatever else you can think of that would STRONGLY suggest that the modification would result in a much more potent compound... only to discover that in man, it was no more potent. I speak of oxymorphone and N-phenylethylnoroxymorphone in this case. So while uncertain, those digitized images in the link are sort of interesting.

But you still have to obtain your norlevorphanol from somewhere.

While people have certainly improved on the original Grewe et al. synthesis, I would suggest that the KEY improvement is that chiral catalyst I mentioned. If it's possible to produce enantiopure dextromethorphan, it's equally possible to use a chiral catalyst to produce levomethorphan.


As you can see from the above, the intermediate is resolved but I'm sure I've read earlier works that used small molecule chiral catalysts.

But even so, unless those potent derivatives I've linked to are as potent in man, it isn't a great target. I mean, you can quite legally obtain the immediate precursor to things more potent than levorphanol. Then I guess you could also play with the N-substituants.
Hi 4DQSAR. Those are very interesting observations that you posted. It is something I need to reflect on. I will also read the papers you have linked as soon as I find the time (especially the one about the asymmetric synthesis). You seem to know a lot about chemistry. More than I do. I wish you were in Germany because I would love to work with people like you and cooperatively create something potentially great. It sucks to do these things alone and not get any live input from fellow chemists. Might I ask from which country you hail? Hmmm, lemme guess...you be an Aussie? Profile pic gives me an Aussie vibe somehow haha.

Anyway, went off on a little tangent there. Thanks again for the papers, your comment and have a nice day wherever you are.


Here the synthesis is here:
Hey Smyth2. I'm already aware of that synthesis but please pay attention to what the text says in the very first sentence: "ONE of the syntheses [...]"
This does not mean that the synthesis as described in wiki is the route of choice. Aside from a dramatic inter-individual yield variance which impacts the cost efficiency from route to route (trust me when I tell you that this is by far the MOST important factor when it comes to a pharma company's decision on which route to synthesize the drug in question), the reagents themselves determine the route of choice by a secondary degree. Since I don't know the yield of the wiki synthesis I cannot say with any confidence whether that synthesis is a more sound choice for manufacture, but my past experience in the industry would give a negative to that (yes past, so maybe that changed in recent years? I'm open to being corrected on that one if someone has any information). I can tell you what I often like to do if yield is unknown to me. I try to make an educated guesstimate of the yields based on similar reactions I know of for all of the routes I want to compare, after which I then proceed to check the prices of the individual reagents used in those routes, add them all up, compare them with the total prices of those individual routes, extrapolate those figures to the assumed theoretical yield, make comparisons to the current market prices of the final products in question and then finally see where the profit margin is the greatest. So if we end up with three ratios, 0.3, 0.6 and 0.15, then the synthesis route that gives 60% profit margin is very likely gonna be the industry-standard route.

P.S.: this ofc totally ignores the type, quantity, solubility and toxicity of all the byproducts which are also a huge factor in the choice of synthesis. A good example is actually the Levorphanol patent I linked. The inventor Joseph P. Haar mentioned that his novel route solves the issue of the hitherto industry-standard route that produced byproducts that could partially not be removed from the final product, which, if you think about the implications of it, is actually a horrifying admission. If I haven't misunderstood what he said, then two pharmacomps might be selling you both the same preparation, but if they used different manufacturing methods for the production, one ending up with residual mercury sulphide (let's hypothetically assume elemental mercury had to be used for some reaction) which is notoriously insoluble in water and is acid and base resistant and therefore, once in the body, won't ever leave the system, then you as the consumer will take those toxins into your body, even though the package insert gives you the wrong impression that the preparation is, irrespective of manufacturer, a pure product. That depends very much on the synthesis used though. I assume this is possible due to some legal loophole?
So all in all the byproduct profile of the route and its cost efficiency are the main determinants which synthesis a manufacturer uses.

DXM is dextrorotatory.
I know, which is why I already wrote twice in this thread that the whole "levorotatory" thing that I mindlessly dropped earlier in regards to DXM was based on a confusion (false association with a closely related compound).
 
Last edited:
A few years ago I discussed synthetic opioids with a Russian chemist. They had been producing methadone for which there was a sizable black market in their nation. They also made dipipanone and phenadoxone to figure out the most 'euphoric' example of the class.

I mentioned levorphanol and there response was simply that although yes, it did appear that it was a deirable target, BUT it's potency meant that it wasn't a facile target i.e. production was too costly for it to be worth all the effort.

That is why I suggested you take a look at some other compounds. Because there are a few things that provide for a more potent product and which require much less work. I GUESS if the 2-(furan-2-yl)-ethyl derivative of levorphanol really is x480M, it would be facile, but still a lot more work than say fentanyl.

As a rule-of-thumb, if a synthesis requires more than three steps, you are going to run into other issues such as solvent management. Or I should say, that was my experience. Yes, for something very potent maybe four, but the amount of effort involved just gets too much.
 
Last edited:
Grewe managed to synthesize (+-)-N-methylmorphinane through an interesting route. The starting compound is Ethyl-2-oxocyclohexanecarboxylate and appears to undergo 5 reaction steps to yield the final racemic product. Depending on the yield it could be a cheaper way to get Levorphanol. Fishersci is actually selling 5G of it relatively cheaply. If yields are good, the DXM conversion route could be seen as a waste of money. I'm gonna have to read the paper to see what his yield was and update it here. I think it would also be wise to put this information on my Levorphanol blog post, so people know about these alternatives. It appears that cyclohexanone derivatives are often the compounds of choice when it comes to the synthesis of Morphinans, which makes sense as they are reactive enough and perfect for cyclization reactions.



I'm currently done with the Levorphanol topic though. The next project will be either the glucose derivative of morphine that I will definitely bioassay on myself to find out if the absence of tolerance translates to humans as well, or I will be creating one of the 14-Hydroxycodeinone derivatives listed by Parfitt and casey...or perhaps even Hydroheroin if I get my hands on some Morphinone. This all depends on the ease of acquisition of the starting reagents, price, yield, complexity of the synthesis itself, etc.
Future syntheses have to wait though because I'm currently saving money to replace my lab devices with high-end Heidolph products. Borrowed that company's rotary evaporator from my friend last week and have fallen in love with that company ever since.

As a rule-of-thumb, if a synthesis requires more than three steps, you are going to run into other issues such as solvent management.
And not just that. A much more annoying issue in my opinion is that syntheses with lots of steps usually lead to a mind-boggling amount of byproducts of which you can easily lose sight if you are not careful. And at a certain point doing more recrystallizations to purify the intermediate/product will simply cause too much loss of yield.
I recently read the old classic "Recreational Drugs" by Professor Buzz and was surprised to find out that Thebaine actually used to be freely available in the late 80s when the book was published. If that was still the case, everyone would be able to cook up their own oxycodone/oxymorphone supply as it just takes two to three very simple steps with very clean results. Those must have been good times for a drug chemist...
 
Last edited:
Janssen stated that he gave up researching the 3,3-diphenylheptanone class of opioids because the chemistry became too complex. He concentrated on the phenylpipeiridine class first discovering phenoperidine and then fentanyl.

When JANSSEN considers it to be too much work, it's likely too much work.
 
Janssen stated that he gave up researching the 3,3-diphenylheptanone class of opioids because the chemistry became too complex. He concentrated on the phenylpipeiridine class first discovering phenoperidine and then fentanyl.

When JANSSEN considers it to be too much work, it's likely too much work.
These aren't even good drugs to start fucking around with. If you could Synthesize Benzos into having a opioid high you'd win a Nobel peace prize
 
If a compound produces an 'opiate high' then by definition, it will be classed as an opioid.

If you mean an opioid with a TI as large as a typical benzo, they exist.

I would also contend that benzodiazepine dependence can be every bit as bad as opioid dependence, so not maybe such a great idea to have a compound produce both.
 
Bro I've never took any drug that felt like Alprazolam & Oxycotonin combined in 1 & I've took alot. (A benzo/opiate pill) would need to be assigned a new gender also
Any drug can create a dependence if you feen enough for it so that's irrelevant.
 
Bro I've never took any drug that felt like Alprazolam & Oxycotonin combined in 1 & I've took alot. (A benzo/opiate pill) would need to be assigned a new gender also
Any drug can create a dependence if you feen enough for it so that's irrelevant.

I assume you mean 'addiction' because it's accepted that many classes of psychoactive don't produce physical dependence.

But it's my opinion that addiction can be harder to deal with than any physical dependence, so I'm not disagreeing, merely checking we are referring to the same issue.
 
Top