Sciencemadness Discussion Board - Warfarin syntheses

Warfarin syntheses

DocX - 2-4-2023 at 12:37

Background
A couple of years ago, an EU directive imposed licensing on all use of pesticides, among them rat poisons. In my country, this resulted in one dominant pest control company almost holding a monopoly on rat elimination for businesses as well as for private individuals.

And what do companies do whenever they're alone on stage? They raise prices. The cost of having someone come over and place little boxes of anti-coagulant poison in the crawl space under your house soon doubled, then tripled. It now costs well over €1000 for a first visit, and if you want the pest control to continue, you're forced into buying a subscription to their services ... you get the idea. Net cost can land you well above €2000.

This calls for rebellion.

The most widely used rat poison in my country is warfarin, an anti-coagulant also widely used by humans as an anti thrombotic medicine for a wide variety of medical conditions. A blood thinner. Ii works by blocking the recirculation of vitamin K in the body (vit K is needed as a cofactor for certain key enzymatic pathways in the coagulative activity in the blood).

It's used as a rat poison (among many other substances in the same family of molecules, 4-hydroxycoumarin derivatives) because of its delayed but definitive action. If the rat drops dead next to, or too close to the poison, other rats are smart enough to avoid it. But the internal bleedings caused by warfarin will not be fatal until a day or two later.
It's an attractive solution. Any accidental poisoning of pets, children or neighbours can easily be reversed by vitamin K injections, and managing warfarin intoxications is routine business for health care givers. Not long ago every other Little Old Lady was medicated with warfarin for their atrial fibrillations. And oh did they overdose.

So, warfarin. Let's go!

Aim of this project
To find a realistic way for the Mad Home Scientist to synthesise his/her own warfarin for rat control using reagents that are cheap and available.

introduction
First off, a search of this site.
Wiki article: https://www.sciencemadness.org/smwiki/index.php/Warfarin
Doesn't offer much in ways of guidance that one cannot get from regular old wikipedia.
I also found this A thread describing the synthesis of coumarin by Fery, but unfortunately using the difficult precursor acetic anhydride. Warfarin is a coumarin derivative, but it seems better to start by directly synthesising 4- hydroxycoumarin, which is a more practical precursor. Warfarin can be produced in a one-step reaction from that:

(Kleeman-Engels "Pharmaceutical Substances", 4th ed)

And then this thread, where ClearlyNotAtara outlines a synth from Aspirin and malonic acid, casually passing 4-hydroxycoumarin on the way:

(https://www.sciencemadness.org/whisper/viewthread.php?tid=16...)

Quote:

acetylsalicylic acid + NaOEt + EtOH + ∆ >> sodium acetylsalicylate + EtOH (g) i.e. just make the sodium salt alkoxide is used here or use something else

sodium acetylsalicylate + EtBr >> NaBr + ethyl acetylsalicylate

ethyl acetylsalicylate + NaOEt (dry) >> 4-hydroxycoumarin (2,4-dioxodihydrobenzopyran) [1]

To begin from something as accessible as acetylsalicylic acid is a very tempting approach, but I must admit I don't quite follow the reasoning behind the route, especially not after reviewing the sources:
[1] http://www.orgsyn.org/demo.aspx?prep=cv1p0235
[2]http://orgsyn.org/demo.aspx?prep=CV1P0149
[3]https://sci-hub.st/10.1016/j.saa.2006.08.023

And that, I'm afraid, is all I'm able to find on this site.

Moving outside, I had little luck with OrgSyn both for warfarin and 4-hydroxycoumarin.
Did find a source for a route from phenol (https://www.sciencedirect.com/science/article/pii/S187853521...):

Quote:

5.2.2. Using phenol
Heating of phenol with malonic acid in phosphorus oxychloride containing twofold amount of anhydrous zinc chloride yielded 4-hydroxycoumarin 1 (Naveen et al., 2006) (Scheme 4).

but it kind of falls flat due to the use of something as inaccessible as phosphorus oxychloride. Several syntheses were also excluded for being too complicated, using really exotic precursors, or just reagents that are illegal or hard to come by for the home scientist. In this group I unfortunately must include acetic anhydride, since it is on the top-tier list of narcotics precursors in all of EU. No simple rat murderer wants to be suspected of cooking heroin. That rules out elegant solutions like this:
https://www.youtube.com/watch?v=oDakl9zxwE4
Youtube is littered with Indian guys scribbling more or less the same synthesis on whiteboards and notebook pages.

Materials and methods
Since I still haven't been able to find a feasible route for this synth using realistic reagents, I need community help. The search ability of people on this site always amazes me.
I've yet to find a path from the rather easily made hymecromone (7-hydroxy-4-methylcoumarin: https://www.youtube.com/watch?v=1OmL1odquqo to anything useful, for example.
Also, I need help thinking outside my warfarin box. Is there another good rat poison that is easier to cook up? Any suggestions are much appreciated.

[Edited on 20232323/4/2 by DocX]

Texium - 2-4-2023 at 12:47

Quote: Originally posted by DocX

And then this thread, where ClearlyNotAtara outlines a synth from Aspirin and malonic acid, casually passing 4-hydroxycoumarin on the way:

Quote:

Could you please post the link to this?

SplendidAcylation - 2-4-2023 at 13:17

Quote: Originally posted by DocX

Background
But the internal bleedings caused by warfarin will not be fatal until a day or two later.
It's an attractive solution.

Jesus.
Poor things, I think I'll sit this one out...

Lionel Spanner - 2-4-2023 at 14:17

The Pechmann synthesis for coumarins uses conc. sulphuric acid and aluminium chloride, rather than phosphoryl chloride - using malonic acid as the beta-carbonyl compound should in theory give you 4-hydroxycoumarin.

DocX - 2-4-2023 at 14:29

Quote: Originally posted by Texium

Could you please post the link to this?

Yes, sorry, it was meant to have been linked in the word "this". I added it as a clear source link.

DocX - 2-4-2023 at 14:34

Quote: Originally posted by SplendidAcylation

Jesus.
Poor things, I think I'll sit this one out...

I think it's probably the kindest rat-fighting method there is. "Internal bleedings" has a dramatic ring to it, but they're actually not painful and usually occur in the gastrointestinal tract, in rats as in humans. You die from painless blood loss, which means you get drowsy and then loose consciousness. Sure beats rat traps.

As for killing rats in any way: the fight between us and them has been ongoing since ... I guess since for as long as we both have been around? Funny thing is, the amount of rats in an area increases almost linearly with the amount of humans. From what I've heard, they're almost always twice as many as we are. So in a way, they're like livestock or pets: only exists because we do. Too bad they show their gratitude by munching away on the wires in the main fuse board ...
I just want an acre where my kids can sleep safely without the risk of getting burnt alive. Internal bleedings or not.

[Edited on 20232323/4/2 by DocX]

Texium - 2-4-2023 at 15:04

Thanks. It looks like what atara proposed is a clever intramolecular Claisen condensation of ethyl salicylate using sodium ethoxide. Like so:

I did a Sci-Finder search and it has been done before, but only in a Chinese patent with 20% yield, using the methyl analog rather than ethyl.
https://worldwide.espacenet.com/patent/search/family/0050314...

The more popular routes all appear to go via 2'-hydroxyacetophenone, for which there are numerous methods.

(This is what me procrastinating on an important presentation for school looks like)

DocX - 2-4-2023 at 17:27

Quote: Originally posted by Lionel Spanner

The Pechmann synthesis for coumarins uses conc. sulphuric acid and aluminium chloride, rather than phosphoryl chloride - using malonic acid as the beta-carbonyl compound should in theory give you 4-hydroxycoumarin.

Yes! Thank you! I instantly latch onto that.

Phenol + malonic acid in a solution of aluminium chloride and ... yes, what solvent should there be? Should there be heat? How much? For how long? Oh those devilish details.

I find this reference: http://www.chm.bris.ac.uk/motm/warfarin/preparation.html, which doesn't offer much in terms of details.
But then there's this one: https://pubs.acs.org/doi/pdf/10.1021/acsomega.9b00257
giving us gems like this:

So, according to that; no solvent, 3 h at 130 C, reported yield around 80-90%.
And from what I gather, an 1:1:1 molar ratio of reagents. Or wait, am I reading this right? That's not AlCl3 they have there, it's some exotic catalyst in a 5 molar ratio... please help.

And then there's this source: https://www.jpsr.pharmainfo.in/Documents/Volumes/vol11issue0...
giving us this reaction:

But that's not 4-hydroxycoumarin as end product now is it? It's some methylated version. And that's not malonic acid as a precursor either, I don't even know what that is... what even is that?
Is H2SO4 as a catalyst still applicable if phenol + malonic acid is used, and will the reaction then produce the 4-hydroxy version?
If "yes", that would mean we could just mix phenol, malonic acid and conc. sulphuric acid in a beaker, let it sit in room temperature overnight and then, just maybe, retrieve an 85% yield of 4-OH-coumarin in the morning, as per this figure from the same source:

And then proceed to isolate that from the surrounding phenolic, acidic goo by some steps unknown to me ... God, I really need an experimental to write one. But if we do manage to isolate it, we can react it with benzal acetone in pyridine to yield ... rat poison!

Not so many OTC reagents there tho.
Phenol I'm sure can be found on Ebay (I know this because I did so)
Pyridine can be made reasonably easy by destructive decarboxylation of niacin (same there).
Malonic acid ... Ebay, apparently.
Aluminium chloride ... nah. But I have to look into the displacement reaction between CuCl2 and aluminium powder. I have some of both, and they're not difficult to get. Or just react some foil with HCl. Or chlorine gas. But should it be anhydrous or hydrated?

But then again... there's this:
https://eng.hekserij.nl/shop/fragrance-raw-material/search-b...
I hate it when that happens.

Still, there's one step left, and we still need to have some
Benzal acetone, and that's not Hardware Store stock exactly. But acetone is. And sodium hydroxide. And benzaldehyde is achievable if you google really really hard ... and then there's this: http://orgsyn.org/demo.aspx?prep=CV1P0077.

It's starting to take form. But for A Bear Of Little Brain like me, much reading still needs to be done.

[Edited on 20232323/4/3 by DocX]

[Edited on 20232323/4/3 by DocX]

arkoma - 2-4-2023 at 17:35

I've had to disconnect SEVERAL circuits running to the far end of my house due to the damn rats chewing on the wires in the attic. To get power to that part of the house I now have a piece of romex running down my hall up against the ceiling where I can SEE it and the rats can't chew at their leisure.

Good Luck in the age old battle DocX!

Fery - 2-4-2023 at 20:37

Warfarin is slow. It lasts days until the coagulation protein level falls down. It blocks synthesis of one of these proteins in liver. Half time of this protein is something like 1 week. Antidotum is vitamin K (eat a lot of tomatoes, cabbage, lettuce etc when working with warfarin). In case of severe bleeding there is no time to wait for synthesis of these proteins and human plasma is administered to the patient (which includes this ready protein from donor people). The usual doses of warfarin administered to people are something like 2-5 mg per day. This dose is therapeutic, not lethal.
Rats are allegedly clever. When finding new source of food, only very few of them feed on this newly found food, the rest of the group waits and start to feed on new food only after some time passes. Here the advantage of warfarin. If using fast poison like metal phosphides (which hydrolyze with HCl in stomach to gaseous PH3) the rest of the group may see the new food is not good to eat.
There could be also predators feeding on rats, if any predator eats warfarin poisoned rat the predator could also die.

[Edited on 3-4-2023 by Fery]

DocX - 2-4-2023 at 22:46

Quote: Originally posted by Fery

Warfarin is slow. It lasts days until the coagulation protein level falls down. It blocks synthesis of one of these proteins in liver. Half time of this protein is something like 1 week. Antidotum is vitamin K (eat a lot of tomatoes, cabbage, lettuce etc when working with warfarin). In case of severe bleeding there is no time to wait for synthesis of these proteins and human plasma is administered to the patient (which includes this ready protein from donor people). The usual doses of warfarin administered to people are something like 2-5 mg per day. This dose is therapeutic, not lethal.
Rats are allegedly clever. When finding new source of food, only very few of them feed on this newly found food, the rest of the group waits and start to feed on new food only after some time passes. Here the advantage of warfarin. If using fast poison like metal phosphides (which hydrolyze with HCl in stomach to gaseous PH3) the rest of the group may see the new food is not good to eat.
There could be also predators feeding on rats, if any predator eats warfarin poisoned rat the predator could also die.

[Edited on 3-4-2023 by Fery]

Not THAT slow. When you put a patient on warfarin, you use the doses you say, and then it takes 1-2 weeks to reach stable plasma concentrations of therapeutic levels. But, also like you say, those doses are tiny- because the therapeutic range of warfarin is really narrow. You measure it by secondary endpoints- one part of the wonderfully intricate blood clotting cascade of co-dependent enzymatic events, the prothrombin complex.
But that measurement can actually shift daily, and there was, and is, a costly struggle to keep up with changing old kidneys and livers of patients taking the medication. That's why health care personnel got so used to dealing with warfarin intoxications: a dehydrated old lady who had a fall, possibly a minor head trauma, and was found to have blood clotting levels through the roof - it was an almost daily occurrence at the ER. Dangerous, very costly events.
That's why warfarin for the most part has been phased out in favour of the so called NOAC (new oral anticoagulants) medicines, which have much more stable pharmacodynamics and wider therapeutic ranges.

But the rats are given HUMONGOUS doses compared to. little old ladies. it shouldn't take them more than a day or so to die.

But how about the chemistry? If one should go with Pechmann, would H2SO4 be feasible?

DocX - 3-4-2023 at 11:40

Actually... reviewing my own digging, I see an error in my thinking.
If I was to skip the synthesis of 4-hydroxycoumarin from phenol and instead buy the easily available Coumarin - is there actually an easy way to convert that to 4-hydroxycoumarin?

So the route would be
COUMARIN --[unknown reaction]--> 4-HYDROXYCOUMARIN --[benzal acetone, pyridine]--> WARFARIN

[Edited on 20232323/4/3 by DocX]

[Edited on 20232323/4/3 by DocX]

Texium - 3-4-2023 at 12:12

Unfortunately I think you're actually taking a step in the wrong direction. I can't think of any conditions that would affect that oxidation selectively, and SciFinder doesn't bring anything up.

Lionel Spanner - 4-4-2023 at 06:11

Found this 1943 paper, which details the preparation of 4-hydroxycoumarins from methyl acetylsalicylate (i.e. the methyl ester of aspirin.)
https://pubs.acs.org/doi/pdf/10.1021/ja01252a007

DocX - 4-4-2023 at 12:01

Quote: Originally posted by Lionel Spanner

Found this 1943 paper, which details the preparation of 4-hydroxycoumarins from methyl acetylsalicylate (i.e. the methyl ester of aspirin.)
https://pubs.acs.org/doi/pdf/10.1021/ja01252a007

Amazing find! Alas, no access. I'll have to fix that.

[Edited on 20232323/4/4 by DocX]

Tsjerk - 5-4-2023 at 03:09

Quote: Originally posted by DocX

Amazing find! Alas, no access. I'll have to fix that.

[Edited on 20232323/4/4 by DocX]

https://sci-hub.se/https://doi.org/10.1021/ja01252a007

unionised - 5-4-2023 at 04:04

Quote: Originally posted by DocX

Background
It now costs well over €1000 for a first visit, and if you want the pest control to continue, you're forced into buying a subscription to their services ... you get the idea. Net cost can land you well above €2000.

This calls for rebellion.

[Edited on 20232323/4/2 by DocX]

...Or a cat.

Diachrynic - 5-4-2023 at 10:53

Quote: Originally posted by Texium

I did a Sci-Finder search and it has been done before, but only in a Chinese patent with 20% yield, using the methyl analog rather than ethyl.

Done in a German paper from 1915 from methyl acetylsalicylicate in 55% yield: H. Pauly, K. Lockemann, Berichte der deutschen chemischen Gesellschaft, 48, 1, p. 31, https://doi.org/10.1002/cber.19150480107

I'll translate the relevant section here:

Quote:

100 g of pure acetylsalicylic acid methyl ester are heated in an open flask in an oil bath to 165 °C. 12 g of sodium metal is added in thin slices, at a rate similar to their dissolution. The pieces should be submerged immediately and distributed quickly, and the temperature should not rise above 175, but also not drop below 160 °C. The precipitation of a brown colored sodium compound begins instantly, and at the end the whole mass is solid and dry. At this moment, adding the last bits of sodium are prone to overheating and finally charring, but this can be easily prevented by cooling the flask in cold paraffin. The hard viscous mass is ground under high boiling petrol ether and after its removal is added to mineral acid, where the formed benzotetronic acid [oxycoumarin] soon precipitates in plate like crystals. After recrystallization from 70% ethanol the melting point was correctly 206 °C and free of salicylic acid. Obtained 43-44 g.

Texium - 5-4-2023 at 12:44

Sheesh, not sure how I managed to miss those papers. I had done a structure search and normally that will pull up everything, regardless of language.

Edit: probably because I was searching for routes using alkoxides rather than sodium metal. I had seen some routes that used sodium metal, but had assumed that wasn’t accessible enough for OP’s needs.

[Edited on 4-5-2023 by Texium]

DocX - 5-4-2023 at 12:50

Quote: Originally posted by unionised

Quote: Originally posted by DocX

Background

This calls for rebellion.

...Or a cat.

Great idea! What is the best route from a base polycarbonate to a full cat? Can't find it on SciFinder.

DocX - 5-4-2023 at 12:55

Quote: Originally posted by Diachrynic

Quote: Originally posted by Texium

I did a Sci-Finder search and it has been done before, but only in a Chinese patent with 20% yield, using the methyl analog rather than ethyl.

Quote:

This seems seductively doable. But also prone to errors. Charring? Those dry synths at high temps sort of scare me. I see pictures of charred and destroyed RBF:s flash before my eyes.

Also, this version is actually tried and mentioned in the paper referred to by Lionel Spanner above, and found wanting:

Quote:

Pauly and Lockemann" synthesized 4-hydroxy- coumarin from methyl acetylsalicylate by adding metallic sodium to the molten ester. They re- ported a yield of 55%. However, by following their conditions we were not able, in spite of many trials, to obtain pure 4-hydroxycoumarin in yields above 13%.

[Edited on 20232323/4/5 by DocX]

[Edited on 20232323/4/5 by DocX]

Diachrynic - 6-4-2023 at 00:28

Quote: Originally posted by DocX

Also, this version is actually tried and mentioned in the paper referred to by Lionel Spanner above, and found wanting

That's a shame. Thanks, that's important information.

DocX - 7-4-2023 at 03:45

Quote:

Unfortunately I think you're actually taking a step in the wrong direction. I can't think of any conditions that would affect that oxidation selectively, and SciFinder doesn't bring anything up.

Yeah, I can see that. Except for biosynthesis that is, but those are messy.
My initial hunch to go for the 4-hydroxy version directly is probably the way to go.

I had an answer written to your post above about 2'-hydroxyacetophenone- routes too, but for some reason I get an error whenever I try to post it. At least one of those routes seems perfectly doable and well describe, but I can't find an accessible source of the starting material.
Whenever I try to answer that post with a link to sci-hub about a specific synthesis I get the error message:

Quote:

The system has failed to process your request. If you're an administrator, please set the DEBUG flag to true in config.php.

[Edited on 20232323/4/7 by DocX]

DocX - 7-4-2023 at 03:53

Trying again (the thing that brought on the error seems to have been a character in the quote, a degree symbol after 100 in the quoted text from the paper):

Quote: Originally posted by Texium

The more popular routes all appear to go via 2'-hydroxyacetophenone, for which there are numerous methods.

Ok! New to me. Found this:
https://sci-hub.st/10.1081/scc-100104003

Quote:

General Procedure for the Preparation of 4-Hydroxycoumarin (3a)
To a stirred suspension at room temperature sodium hydride (60% in mineral oil, 60mmol) in anhydrous toluene (30mL) was added dropwise a solution of 2-hydroxyacetophenone (12mmol) and acylating reagent (18mmol) in anhydrous toluene (6mL). The reaction mixture was heated at 100C for 3h. After cooling, the solvent was evaporated and the residue was treated with water (10 mL). The resulting solid was filtered, washed with water, and then recrystallized from methanol or ethanol to give 4-hydroxy- coumarin as a pale yellow crystal.

But how oh how does one get access to dat 2'-hydroxyacetophenone?

[Edited on 20232323/4/7 by DocX]

[Edited on 20232323/4/7 by DocX]

Mateo_swe - 26-4-2023 at 04:50

Poor rats, they are living animals too.
If you need to kill them, do it quickly.
Just put a big tub filled with water then a thin round pole smeared with lubricant over its opening at the top.
Put a piece of food at the center of the pole.
The rat or mice will go out on the pole and loose grip falling inte the water and drown.
Or if you use a tub with high walls you can catch them alive and drive them out in the forrest somewhere, thats seems like the friendliest thing.

Better than rats to go around like zombies for days or even weeks, probably not plesant.
I have seen what Warfarin does to rats, it doesnt seem nice at all, there must be better ways.
A cruching trap would be best but if you have a big rat infestation thats going to be hard to get them all with such trap.

Regarding the chemistry, intresting..

DocX - 26-4-2023 at 11:47

Quote: Originally posted by Mateo_swe

I have seen what Warfarin does to rats, it doesnt seem nice at all, there must be better ways.

Really? Pray tell, how and when did you see this?
A warfarin overdose should most likely result in gastrointestinal bleeding. The gut of mammals has a HUGE, extremely vasculated surface area, and bleeding disorders generally start there. To sieve blood into the gut is a pretty painless thing, you get tired, drowsy, shot of breath and faint.

Or it's a haemorrhage stroke. That can be painful, but not for long. Not like drowning.
Bleeding is sort of painless. So I really wonder what you saw?

And regarding the chemistry I have actually reached a dead end when it comes to the reagents. My hope was there would be a route with affordable and accessible precursors, but I still haven't found one.

[Edited on 20232323/4/26 by DocX]

Fery - 28-5-2023 at 12:47

Reviving the thread. Did anyone succeed in 4-hydroxycoumarin synthesis?
Here I found a German page with coumarin synthesis:
http://forum.lambdasyn.org/index.php/topic,1471.0.html
I synthesized the second reactant (benzylideneacetone) here:
https://www.sciencemadness.org/whisper/viewthread.php?tid=15...

Fery - 10-10-2023 at 06:27

This seems to be a suitable home lab way for warfarin. But only racemic product. For obtaining stereoisomers it would be necessary to have (R,R)- or (S,S)-1,2-diphenylethylenediamine:
https://pubs.acs.org/doi/10.1021/ed800040m

Benzalacetone synthesis is here:
https://www.sciencemadness.org/whisper/viewthread.php?tid=15...

The 4-hydroxycoumarin may be obtained from phenol and meldrum acid:

Quote:

Treatment of phenol with meldrum acid under solvent-free condition at 90°C afforded 3-oxo-3-phenoxypropanoic acid in 92% isolated yield, that transformed to 4-hydroxycoumarin in 75 % and 48% isolated yield, upon treatment with Eaton’s reagent or polyphosphoric acid (PPA)

Attachment: j.arabjc.2015.06.012.pdf (1.6MB)
This file has been downloaded 233 times

The meldrum acid may be obtained by reacting malonic acid + acetone + acetic anhydride + cat. H2SO4

Attachment: CT9089300598.pdf (244kB)
This file has been downloaded 184 times