Hi!

I am making small scale tinctures and extracts but have been running into capacity problems lately. Because applying heat to many herbs degrades some constituents and impacts the flavor, cold maceration has so far been the preferred choice. But as volume increases, dealing with larger amounts of liquid becomes an issue. Maceration needs a lot of solvent. Also, buying larger amounts of ethanol is prohibitive, and there is always the problem of solvent recovery. A soxhlet takes care of that to a certain degree, however, the glass soxhlets are not commonly made larger than 2-3L.

It seems like the only commercially available option for this type of situation is the Coldfinger from Eden which can handle vacuum and up to 50L capacity - but it costs 14'000USD. So I am planning a DIY project to build a vacuum soxhlet based on a 20L pressure cooker. But before going for it I want to make sure there isn't something obvious I'm missing - or if there's another way to go that's better in the first place.

The basic idea is to re-purpose a pressure cooker into a vacuum chamber. Then, below the lid a circularly winded copper tube is attached as a condenser. Two additional holes are drilled into the lid for the cooling water circuit. The main pressure valve that is already built into the lid is used as the vacuum valve. Add a pressure gauge and essentially that's it - herb is added in a suspended basket, ethanol added, and vacuum applied. At a pressure low enough, the ethanol will boil at room temperature and on goes the soxhlet. Like with the Eden design, the same apparatus could then be re-fitted to work as a solvent recovery system in the very same fashion except that for the herb basket there is a funnel which collects the condensate and transfers it to a second vessel.

Here is some documentation about the Eden apparatus: https://edenlabs.com/med...iller%20Brochure2013.pdf

So, together with a steel smith that doesn't sound like rocket science right? - am I overlooking a crucial point? Edenlabs sure has done its research, so my concern is that a DIY apparatus is not working properly because of some obscure ratio or formula that wasn't thought of.

And the more broad question - does this seem a sensible next step to scale up or are there smarter solutions for continuous solvent extractions that I'm missing? Would a stirred reactor approach (as described in the brochure above) be more efficient?

Thanks everyone for ideas!

Could work, you'd need a vacuum of 60mmHg for ethanol to boil at rt (25C). You'd want to make sure everything is airtight. You can actually buy a vacume chamber and a pumpmfornwhst you're looking at spending.

Another thing is your vacuum will probably suck out the ethanol vapor. So instead of coils you'd want a reflux column instead. You'd have to cool the snot out of it so the ethanol didn't reboil. Remember, you did just drop the boiling point by about 59C.

I contend an= simpler solution, a metal pot and a concave lid. (Perhaps even a metal cone). Pour out the alcohol in the bottom, insert the basket and fill the cone with ice. The ethanol will condense on thr cone amd drip off at the point going through the basket.

The vacuum sucking out the ethanol vapor is a good point. The idea behind the coils is to increase the surface in which condensation takes place - with a column the reflux is concentrated in a small area of a few centimeters only. That wouldn't re-distribute the condensate over the herb basket equally. But maybe a combination of both would take care of the vapor loss. All vapor that makes it past the coil-condenser goes into the column where it will hit a second condenser.

I think the same problem would be happening with a concave lid - the condensate would aggregate mostly at the lowest point in the center - no?

One thing though - how would I know when the ethanol really is boiling? Without a see-through window only calculation would indicate the boiling point, but I'm never getting accurate results with that. So over-boiling would be an issue, which could lead to foam burst, not good for sure..

If your vacuum pressure is good, there's no need to check. Condensing into a single point would be prime unless you want the vapor to run down the sides. It's a neat idea but it requires quite alot of expenses/labor when an actual soxhlet can be picked up for cheap.

A reflux column would let your vapor rise up, and recondense; raining solvent into your goodies. Also note how all reflux/soxhlet setups are open ended. What you're referring to is more or less a vacuum chamber with a built in condenser.

Then for a thimble just take like a hydrometer tube, drill some holes down at the bottom. Pack in some cotton or glass wool on the top/bottom of your material to be extracted, and have your reflux drip into that. You'll find NOT using a vacuum to work significantly better. It also means better cooling, and better reflux. Which means more alcohol runs through your material. Which means faster and more efficient exteactions.

I'd like the vapor to be even distributed over the entire herb basket. With a single point it would only extract the center of the basket.


A column is definitely a good idea! It will be part of the design, certainly to catch every bit of vapor! For the main condenser I don't see it because of the single point condensation...


RotoVaps are essentially just that too - a vacuum chamber with a condenser. Soxhlets are commonly built open ended but can be built to receive a vacuum.


For sure it would be preferable, but quite a number of herbs just don't respond well to heat. To preserve flavor and full-spectrum constituents heat isn't good. Probably heating to around 40 degrees makes sense, as it makes condensation easier but doesn't yet degrade constituents.

it's more of a distillation app that has a vacuum-take off valve/ purge inlet. it is not a fully-integrated unit, but rather, a modular system of components.

with some models, like the vertically oriented R-series, a soxhlet may be implemented (those models probably have a hot oil bath too)
applying vac in this situation is probably not necessary for what a soxhlet was designed for: using low-boiling solvents.

Methylene Dichloride boils around 40c, and iirc acetone is close by as well.

yea, those don't need vac. even a water bath would boil them.

I have built a small version of what ice suggested.

Used a bodium Assam tea pot with a steeping thing and a margarita glass with the stem cut off filled with ice as the condenser lid. Got the idea from the Coldfinger design but it only cost me one pound to build (for the teapot at a car boot sale) works really well and you can swap the mesh steeping bit for a cup and recover the solvent afterwards.

Don't have vacumme but I only use it for hash extractions and it works fine with ISO at 60 to 70c.

Love the fact that it is a clean solvent dripping through, pulls way more than a soak would and allows a much higher concentration to build up.

You could make a big version for 20L extractions real easy.

One of those stainless steel things for steaming veg could work if you inverted the lid and put ice on top.

I think that once you introduce a vaccum that you will need to build a cold trap to catch all your solvent that boils off in response to the reduced pressure.