automated liquid-liquid extraction

[der-seemann]

An idea came to my mind.
Supposed you have a based liquid, containing some alkaloids you want to extract with a NPS.
And you got pipes, tubes, some more flasks and a pump that is fine with the NPS.
Then you could maybe set up a system where the NPS is pumped into flask 1 and bubbles thru the mother liquid, then flows to flask 2 where any small particles from the mother liquid can settle, then to flask 3 where it is bubbled thru a small amount of vinegar, then to flask 4 with vinegar again, and then in an NPS storage flask. from there the pump will suck the NPS and slowly pumped it into flask 1 with the mother liquid.

Flask 2, 4, and 5 are not really necessary, but should improve the system and the quality.

A peristaltic pump with suitable (silicone? rubber?) hose for 10 - 20 € from ebay should work.

All flasks are closed with a plug where only the pipes go thru. Just flask 5 is open to compensate any pressure that could build up in the otherwise closed system.

So your alkaloids get picked up in flask 1 and will be saltet out in flask 3 and 4.

You can let it run by itself until all alkaloids are extracted and concentrated as acetate in a small amount of clean liquid.

No solvent will be lost, as it is a closed cycle.

The whole setup should be placed in a container to catch any leakage that could happen while operating unattended. For the same reason a harmless NPS should be chosen.

Credits to downwardsfromzero for reflection and suggestions!

What do you think? Any suggestions welcome.

 

[downwardsfromzero]

I've thought about designs along these lines (one with a freeze precipitation unit and a filter) from time to time, then asked myself, "What would be the advantage of going to the trouble of building this?"

I never really came up with an answer beyond, "It might be fun," and considering I don't really have much desire to produce alkaloid extracts on an industrial scale (j/k) it never got further than that. I did build a continuous distillation rig once, though.

I would suggest you try a pilot set up with just water and oil/solvent first. There is also a distillation-based continuous extraction device called a Clevenger apparatus. They're pretty pricey.

And a continuous liquid-liquid extraction apparatus, seemingly without a catchier name:
[Edit: ok, it's not embedding, I'll try attaching the picture.]
Your set up would save you having to buy two of the latter, I guess

 

[der-seemann]

well, i was thinking in the range of 500 ml tea. That's far away from large scale
These nice glass wares partly inspired me to my setup and are probably in the same dimensional range and not meant for processing many liters.

There are also high tech systems for real continuous extraction, suitable for larger scales, but what for?

Of cause, a separation funnel or even the good old turkey baster would work too. But it could be fun figuring this out I thought.

The advantage would be that you can skip the boiling / concentrating of the tea and get a clean, defatted and concentrated alkaloid-acetate-solution

The second a-b ( from the vinegar-solution to final freebase ) has still to be carried out by hand.

 

[downwardsfromzero]

The overall idea I like, don't get me wrong. You would need to think carefully about the pump. An alternative would be to have a boiling flask as the 4th container and the NPS would get returned to the first flask by distillation instead. That might screw up the syphon pressure, of course, unless you put the boiling flask significantly lower than the rest of the apparatus. Then you'd need regulating valves on each of the syphon lines (which is probably sensible anyway). You would need to ensure that no one of the syphon lines runs dry at any time as well.

How would you prime the syphon lines?

Definitely do a few test runs with just water and light oil first, and clamp all the bottles securely. Think carefully what the failure points are. Are they greater in number than what typically might be found during a simple manual extraction? Any one of the syphon lines could come loose and leak NPS. I'd strongly recommend using the apparatus only within a secondary containment unit (HDPE/PP storage box).

You might find it involves a lot of fiddling around requiring constant attention, or computer controlled monitoring and flow regulation if you're that way inclined. The main problem will be ensuring that none of the three syphon lines runs dry, and also the return pumping line if the pump isn't self-priming. That pretty much means controlling the flow on each of the four fluid lines.

I can imagine doing this for fun, but don't expect it to make extraction any easier.

And if you've got that far, surely it gets tempting to come up with something to sort out the acidic phase as well :twisted:

 

[der-seemann]

You would need to think carefully about the pump. An alternative would be to have a boiling flask as the 4th container and the NPS would get returned to the first flask by distillation instead. That might screw up the syphon pressure, of course, unless you put the boiling flask significantly lower than the rest of the apparatus. Then you'd need regulating valves on each of the syphon lines (which is probably sensible anyway). You would need to ensure that no one of the syphon lines runs dry at any time as well.

How would you prime the syphon lines?

the peristaltic aka hose pump should be ideal for this. It isself priming and very fine tuneable, in the range of 0-4 liter /h.

as the flast 1-3 are sealed, its not really syphoning but just an overflow-system. so no prob with initial filling the pipes and controlling the liquid-levels.

you can first fill bottle 1 with motherliquid just one cm below the exit pipe and bottle 3 and 4 with a few cm of vinegar (over the end of the entry-pipe).
then you seal the bottles 1-3.

now you take the exit pipe from bottle 4 and put in a storage container of nps.
start the pump and wait until the nps has filled the bottles 1-3 up to there exit pipes and bottle 4 to 50% or 75%.

then put the exit pipe of the storage container back to bottle 4. now it will run in closed loop...

Definitely do a few test runs with just water and light oil first, and clamp all the bottles securely. Think carefully what the failure points are. Are they greater in number than what typically might be found during a simple manual extraction? Any one of the syphon lines could come loose and leak NPS. I'd strongly recommend using the apparatus only within a secondary containment unit (HDPE/PP storage box).

a leagage container is a good point.
also the choice of nps. if it should run unmonitored, its good to choose a relativly safe not volatile nps, no naphta.
Othervise if you have a leakage the nps can evaporate over night and when cou come back an switch on the light : booooooom!

You might find it involves a lot of fiddling around requiring constant attention, or computer controlled monitoring and flow regulation if you're that way inclined. The main problem will be ensuring that none of the three syphon lines runs dry, and also the return pumping line if the pump isn't self-priming. That pretty much means controlling the flow on each of the four fluid lines.

as stated above, the system should be self regulating and easy to fill and prime.

I can imagine doing this for fun, but don't expect it to make extraction any easier.

once it is set up, it could also open new possibilitys in extracting from very weak but local and sustainable sources.

And if you've got that far, surely it gets tempting to come up with something to sort out the acidic phase as well :twisted:

you think of an automated FASA or freeze-precipation setup?

 

[downwardsfromzero]

You seem to have thought this through well, and your use-case is a very good reason for having a semi-permanent set-up. There would still be some amount of sample preparation involved before it could be loaded into the first extraction flask. So, say you had an acidic phase - presumably that would be evaporated down to a sensible volume? Or would you rather, perhaps, spend the energy on pumping both a larger volume of solvent and having an additional reservoir of crude plant extract that would be pumped through the first flask?

Have you performed any preliminary tests yet? Have you thought how the apparatus would be affected by changing temperature and any ensuing expansion/contraction? Oh, and a continuous liquid-liquid extraction system best has rotating nozzles on the end of the end of the solvent delivery dip tubes. As I get the feeling you already appreciate, this increase the P/NP solvent mixing interface.

It's certainly wise to use a low-hazard solvent system if this would be left running unattended. Perhaps n-dodecane would be worth trying as it's practically hazard-free and I would bet it's a pretty good solvent for DMT (if that's a reasonable assumption to make about your extraction target ). Low-odour lamp oil is completely OTC and broadly similar.

you think of an automated FASA or freeze-precipation setup?

These ideas have been hurled carefully into the magic junk room at the back of my mind and we'll see what comes back out sometime. Already I'm envisioning a parallel set-up for defatting of the initial acidic phase, which could be fairly useful.


All in all, I'd love to see this up and running. Come to think of it, I've a peristaltic pump knocking around in a cupboard somewhere...

 

[der-seemann]

I've extended the image and added a NPS reservoir flask before the pump.

Have you performed any preliminary tests yet?

no, it's ony theoretical so far

Have you thought how the apparatus would be affected by changing temperature and any ensuing expansion/contraction?

The NPS level in the reservoir / storage flask will rise and fall, so no problem!

Oh, and a continuous liquid-liquid extraction system best has rotating nozzles on the end of the end of the solvent delivery dip tubes. As I get the feeling you already appreciate, this increase the P/NP solvent mixing interface.

thats right, but that could end up in an emulsion. So I rather pump very slow, like 3 liters / h and let it run for a day or a week if necessary. A magnet stirring at flask 1 could be good i guess

Already I'm envisioning a parallel set-up for defatting of the initial acidic phase, which could be fairly useful.

Is this necessary? all the fats will end up in your circulating NPS, the acetat solution should stay free of them.

 

[downwardsfromzero]

Fat + base = soap. That's one good reason to defat the acid cook.

Here's a possible good OTC solvent choice: