by Dave Arnold
The Buchi corporation invented rotary evaporation, and rotary evaporators. They design their machines to remove solvents from chemicals in a laboratory setting. Because rotovaps distill gently at low temperature under vacuum, they happen to also be good at preserving flavors — much better than traditional distillation. The rotovap lets you make uncooked port wine reductions and the freshest herb liquors in the world. It helps you remove specific unwanted flavors from any liquid, and keep the ones you want. It’s your aroma thief, removing and retaining volatiles so you can drink them. (For in-depth information, see the rotovap primer.) Buchi visited me several weeks ago to discuss rotovapping in the kitchen. I showed them my in-progress prototype (my long-time back-burner project, as loyal readers know: see here). I’m happy to say that Buchi was receptive to a lot of my ideas.  Rob Crotchfelt, our regional sales rep (and Cooking Issues’ newest best buddy), Herve Lacombe, President of Buchi Corporation, and Matthias Schacht, Product Manager for Rotary Evaporators at the main factory in Switzerland (a real rotovap engineer!) engaged me in a great discussion, then left behind a spanking-new Buchi R210. It’s still built for the lab, but even in the kitchen it is one sweet machine.
The R210
A quick history of my rotovap lineup. About five years ago I built a rudimentary vacuum still, a bootlegvap that was a finicky pain in the rear.  Fed up, I bought an early 80’s EL131 Buchi rotary evaporator on eBay the next year (they’re cheap at auction – check it out). I customized it to better suit my kitchen needs, adding a peristaltic pump to remove distillate as it was produced. My pride and joy, that rotovap pumped out many bottles of delicious booze. It earned its retirement, which was effective the day I got my state-of-the-art R210.
My new Buchi evaporator is much easier to take apart and clean than its predecessor– which means I’ll be much calmer when Habanero pulp sprays all over the inside. It’s sturdier than the EL 131, and it has a motorized lift mechanism to move its glassware up and down – a feature I initially thought was extravagant, but now realize is essential: it keeps you from slipping and breaking another $350 worth of glassware. The new machine leaks a hell of a lot less air, so no more hours testing and fixing vacuum leaks.
My sit-down with Buchi
Good news: Buchi showed interest in developing a more kitchen friendly (read: cheaper) machine. They aren’t really up for something as different as I proposed — a ground-up, purpose-built kitchen tool – but they see the logic in making some chef-and-bartender-friendly modifications to their design. Buchi agrees that the food world could use its own rotovap, because:
1. Kitchens and fragile, expensive lab-quality glass don’t mix
Lab-vaps must withstand a wide range of corrosive chemicals, so they’re made with expensive (and fragile) lab glass.  In the three years I’ve run the tech intern program we’ve destroyed about $1500 worth of lab glass –ouch.  Kitchens don’t require the purity and chemical resistance of glass. We are better off with more durable materials, a fact I demonstrated to the Buchi guys by banging my fist repeatedly on the polycarbonate/stainless-steel skeleton of my prototype (the machine won). The Buchi team was convinced.
Accurate variable speed rotation is another pricey rotovap feature essential for labs but not kitchens. Buchi agreed it could be done away with.
2. Vacuum control is great, but really expensive the way it is done now.
Distillation is boiling. The boiling point of a substance is a combination of a temperature and pressure. In traditional distillation, pressure is constant (atmospheric pressure) and temperature is adjusted to control the distillation. In rotary evaporation, on the other hand, temperature is the constant and pressure is the variable, controlled by the level of vacuum. Old-school rotovap junkies (and lab paupers) fly their machines by adjusting the vacuum pressure with a bleeder valve.  A bleeder valve and an inexpensive refrigeration-style vacuum pump costs less than $120. That system is cheap, but it requires constant monitoring, is loud, and takes lots of practice to master. Modern rotovap vacuum pumps and vacuum control units solve these problems, but are very expensive –several thousands of dollars.
The Pump
Modern lab vacuum systems use quiet, dry vacuum pumps. They are a joy to use and don’t make nasty oil fumes, but they will set you back at least a thousand dollars. We couldn’t think of a cheaper alternative to the loud, oily refrigeration pumps that would both work and be affordable. I guess we’ll deal with noise and fumes.
The Controller
Modern vacuum controls let technicians walk away from the machine while it’s running. These controls are not cheap. Their vacuum sensors are expensive, and so is the software development they require: vacuum gauges measure only the level of vacuum, and require sophisticated microprocessor control to determine the optimal pressure for a distillation based on extrapolations from the pump’s current vacuum-pumping rate.
And adding insult to injury, these pricey automated distillations are not really good for kitchens. Kitchen distillation, unlike most lab distillation, needs to be quick (for flavor, and for practicality) and involve multiple solvents (usually ethanol and water) whose boiling points are constantly changing. The automated lab systems tend not to be speedy enough for kitchen distillations and don’t handle multiple solvents well.
I would love the luxury of walking away from the rotovap while it is running. I have long maintained that good automated control of distillation would be possible using inexpensive thermocouples rather than expensive vacuum gauges. Controlling distillation using the temperature difference between areas with and without distillation vapors in the condenser would ensure an aggressive distillation is maintained throughout the distillation cycle – and it would be cheap to build. I was elated to find that Buchi offers this option on their high-end controller. Unfortunately, it’s only offered as an addition – not an alternative — to their high-end vacuum controller. After I test it, I’ll tell you how it works. Maybe they can make a cheap controller that uses only temperature.
2. Lab rotovaps lack features kitchens need, like:
Ability to pump out the distillate:
To make my rotary evaporator kitchen-worthy I add have added a pump that constantly pumps out distallate. I have gone through several versions of this pump. I eventually settled on a peristaltic pump. Peristaltic pumps work by squeezing liquids or gasses through lengths of tubing. Most tubing won’t work. I tested dozens of varieties and ended up with semi-rigid polyurethane tubing; it’s connected to the rotovap using a spherical glass adapter. On my old EL131 rotovap, the pump was separate– inconvenient when raising and lowering the machine. I bolted the pump on the side of my new rotovap, so the automatic lift raises and lowers it– very swanky. I developed this system for two reasons: 1. you want to taste the distilled product as it’s being made, so you know when to cut distillation off; and 2. You want to remove product as you make it so the distillate doesn’t sit in the machine over the whole run – I believe that doing so kills the flavor.
It’s my hunch that in the unaltered rotovap, flavor compounds are always re-distilling a little bit in the flask, and they aren’t always re-condensing 100% –they get lost through the vacuum pump. Matthias, the engineer, made an interesting counter-observation on this point. “Assume,†he said, “that your rotovap had no leaks in it whatsoever. You would be able to suck an initial vacuum, turn off the pump, and do the entire distillation without turning it back on. In that scenario there could be no flavor loss because the system is sealed.†I had to agree.
In a leak-proof rotovap, flavor won’t be lost in the flask. I now believe that leaky rotovaps and continuously running vacuum pumps are the prime cause of flavor loss in an otherwise properly done distillation. In fact, I have noticed that leaks in different places have different impacts on flavor. Maximum flavor loss occurs when the air leak is located at the distillate collection point. The incoming air strips out the flavor.
The guys from Buchi understood why I added the pump, but I didn’t get the feeling they wanted to go in that direction. They say they currently have a system that, with a little modification, will allow the operator to remove the distillate in small batches for testing without breaking the vacuum —  I need more information before I can weigh in on that one.
Ability to pump out reduced product or waste:
The lab rotovap’s distillation flask is constantly rotating, so you can’t pump the spent product out as you are working. There’s no way to drain from a rotating vacuum flask, so waste or reduced product builds up inside the distillation flask. In our system you can only rotovap about three liters of product before you must break the system down and empty the flask. In my ideal rotovap design, the distillation area doesn’t rotate, so it can be drained with a pump similar to the one I use to pump out distillate. You can operate continuously if you can pump out product, pump out waste, and add fresh product without stopping the machine. If you can develop a good automatic distillation system, you could run continuously and unattended. The Buchi guys weren’t really willing to go down this path with me.
Easily cleaned flasks:
The Buchi guys left me with a distillation flask that I really like: it is built like a wide mouth beaker with a screw-on cap. It is half the capacity of my narrow-necked 3 liter flask, but it is so easy to clean that it makes me smile just thinking about it. Cleaning out the distillation flask is typically a crummy job because it’s very hard to scrub the inside. Additionally, when we make thick reductions it’s hard to recover all of the product from the narrow necked flasks. The beaker flask fixes those problems. If only they made a three liter beaker…
No boil-over:
One of the big problems with the rotovap is boil-over –you start a distillation, and – blammo- the trapped air inside your product foams up and sprays all over the inside of the rotovap. Big drag. I talked with Matthias about using a fritted material in the distillation path to prevent bubbles from getting into the condenser side of the machine. He said Buchi makes something similar, but it is problematic because the fritted material restricts too much airflow. The vapor duct that separates the distillation side from the condensing side is fairly narrow, and the speed of the vapor shooting through it is quite high. Restricting airflow through that passage isn’t such a great idea –which is why they nixed another of my anti-boilover ideas: an electrical sensor that tests for boiling liquid. They make those sensors for their larger units, where constriction at the vapor duct isn’t so severe.
Happily, the beaker flask Buchi dropped off has made possible a cheap and simple boilover remedy –Scothbrite pads. I bought some fine pads (the white ones) and cut them to fit just inside the head of the beaker. It barely impedes the flow of vapor and is difficult to clog (although I have managed to do it.) But you must be careful not to let the pad fibers get on the sealing surface of the beaker.
Next up in the rotovap chronicles:
What I learned in England; some experiments with mint; a refreshing summer drink.
Why have the flask rotate to increase surface and not just a mixer (similar to the elastic ice-cream device you featured earlier)? Would make your prototype maybe also more robust… and you could add 2 more peristaltic pumps in the heating chamber to let in new liquids and remove leftovers since the chamber doesn’t have to move anymore.
Maybe even this “mixer” can be heated (think of a rotating spiral from those condensers, just with tempered water!) to allow for better heat distribution (to heat 3l of liquid in your prototype to the same temperature might get difficult)
Also you could add a “trap” (http://en.wikipedia.org/wiki/Trap_%28plumbing%29) to catch minor boilovers.
Here’s a small paint-doodle of a design with stationary flask + trap (including the possibility to reuse “trapped” liquds from boilovers): http://img337.imageshack.us/img337/8236/sukrim1000.jpg
Howdy Sukrim,
My impression is that the thin film you get from rotation is ideal. The image you posted is essential similar to what I’m building except the mixer blades are replaced by a cylinder (to increase surface area) and the shaft doesn’t extend into the chamber (it is linked magnetically). The input can be done without a pump (vacuum suction is sufficient).
Hi Dave, from what you’ve written do you think that there is a real chance that Buchi may be looking at manufacturing your design?
Re: magnetic drives I’ve been searching to find something suitable, lots of things available that would be suited to use on an oil rig or companies that would custom design one but it would be great to know if you have found something readily available at a reasonable price.
Out of desperation I’m now looking at ripping a pond pump to bits.!
I don’t think Buchi is interested in making my design. I think they are interested in modifying their own product to make it more kitchen friendly: simpler, more robust, cheaper materials, etc.
Check out mcmaster.com part number 9199T2. That is what I use. For a similar product see: http://www.huco.com/products.asp?p=true&cat=506
cool article! Old(er) school rotovap junkies use an “aspirator” for vacuum.
http://en.wikipedia.org/wiki/Aspirator
A plastic one is probably <$10.
You would need a constant flow of water (sounds wasteful, I know. In the lab we used industrial water).
The other problem with the aspirator pump is that we are often dealing with vacuum pressures down in the 20 mbar range cause we are using low bath temps and distilling water off of sugar solutions, etc. I once tried to hook a bunch of aspirators up to a flowjet pump in a cooler full of ice water (learned that from a meth lab website). Still couldn’t get it to work too well.
a water aspirator is ideal for low bioling point organics like DCM or EtOAc – but it will never remove water without a significant amount of heating. i figure in the kitchen water is the only solvent you’re really looking to remove. yeeee.
Mostly Water and EtOH
Hi Dave,
Could you suggest any elegant solutions for producing small quantities [liter, 2 liter per 3 months] of mirin for cooking? A compact, ingenious design? Industrial mirin production seems to have condensed the tandem saccharification and fermentation steps of tradition into a single process where koji base and distilled spirit are simultaneously loaded into reaction chambers.
Could this simplified scheme be replicated in terms of fermentation equipment at home? What might be a “least cost, least complexity” set-up that could allow someone to tinker with the canon, e.g. experiment with boiled brown rice instead of steamed Japanese grain? Very small quantities, so that failures do not waste food.
Thank you for your advice.
g
Howdy GT,
I don’t know much about mirin production. However as you say, the work I have read uses koji and distilled alcohol rather than a traditional fermentation. I found an interesting article that analyzes different rice types (Thai rice, cause the paper is from Thailand) and includes a lab scale recipe. I’ll look some more.
http://www.thaiscience.info/Article%20for%20ThaiScience/Article/1/Ts-1%20production%20of%20seasoning%20mirin%20from%20thai%20rice%20by%20fermentation.pdf
I don’t see why there couldn’t exist a vacuum controller that uses only temperature sensors and a simple PID algorithm or maybe even as simple as a lookup table. As long as we’d be talking about a custom vacuum controller, why not an infrared transmitter and receiver to “watch” for a boil over.
I think the easiest way to test for boil-over would be conductivity, because a light based meter could become occluded.
So Dave, are you gonna keep building your design? or just wait for Buchi’s improvements to their R-210??
cheers!
Good to hear from you Patricio,
I am still working on it, just haven’t had time.
Hey Dave I got a couple of questions, since I started reading your blog I have been really interested in all the drinks aromas and all the cool things that you have been doing with a rotovap.
I have approached Buchi here in New Zealand and they are very interested in supplying the restaurant with a demo unit for experiments.
They have suggested me that the best settup would’ve been;
*Rotavapor R-215, Vertical Condenser P+G SJ 29/32, 230V with Display.
*BUC71000 V-700 Vacuum Pump
*BUC47299 V-850 for R-21X & V-7X0 (100V-230V) Incl. support, communication cable
*BUC47170 Woulff Bottle
According to the information that I have received from them with this set-up is the easiest way to get distillation done specially if we get the probe for automatic distillation.
They have also suggested to get the unit with the plastic coating glass parts for extra protection in the kitchen environment.
Is this set-up the ideal or would it be too much for the use that you do?
New Zealand! The promised land of legal distillation. Good to hear from you. That setup sounds right but to do the auto distillation with temperature you need the V855 not the 850. I don’t see the chiller listed in the list. Also, get the 4 liter flask.
Howdy Dave,
I am happy to see that Buchi, one of my favorite lab equipment companies is acctually following your work and trying to help as much as they can! In my industiral career I had many situations where I thought I had a brilliant idea how to improve some lab equipment and always found out how hard it is to convince the marketing guys, let alone the engineers. If cou succeed it tends to take ages, too. Good luck there.
I have one unrelated idea you might want to try: You have many cases where you are only interested in the desillate and not the residue (remember in a chemsitry lab it is 99.9 percent the other way around.). So if you have a foaming problem in the rotating flask you could add an anti foming agent (off course one that is approved for medical uses)
Try Simethicone!
http://en.wikipedia.org/wiki/Simethicone
It will break up foam very efficiently. It is tasteless and none of its two ingredinets (a silicone oil and silica to support it) will still over.
Get some and pop a tablet into your flask to avoid burping (just as in real lifel LOL).
Howdy Schinderhannes,
Good tip. Nastassia is looking into it. On a side note I just started experimenting with a cold finger/ln setup on the rotovap. I hate that I can’t tell how much I’ve distilled, but I’ve been getting some interesting results.
Hi Dave,
it really looks like you are darn close to an ghetto lyo with this on.
You have a new and well sealing rota, you have a cold finger chiller, and you have an awesome flask for food stuff. Only thing I don´t know is how good a pump you have, but maybe the following is worth a try:
Take some clarified juice, lime, rasberry whatever. Put like 4-5 ounce into your flask.
Make a dry ice bath (a bowl large enough to put you flask in, filled with chips of dry ice and some booze.)
Rotate by hand your flask in said bath till you have a even thin layer of frozen juice on the inside. (as cold as possible, take your time.)
Put the flask on the rotavap, wrap it with aluminum foil, a layer of insulation (paper towels) and a second layer of aluminum (leave some place to peek inside….)
Pull as much vakuum as you can with coldfinger full of LN, as small a collection flask as you have (nothing will go in there anyways), no rotation and of course no waterbath.
See if the chilling by sublimation of the ice is sufficient to keep it frozen as you freeze dry.
Will take several hours….
Good luck!
If you succed you will get a layer of very light fluffy foam tasting like rasberry juice.
Put this as decoration on a nice desert and enjoy!
Will try. Do you think dry ice is good enough for the cold finger as well? My pump gets down to 4 mbar.
Dry ice in the cold fingers in plenty sufficient, commercial lyos cool the reciever side to -60 to -70..
4 mbar is prolly borderline but I´d give it a try.
Good luck.
The traditional is dry ice and acetone. Although with your near limitless supply of liquid nitrogen….
Hi Dave, a very nice approach to using the Buchi in the kitchen. The flask must spin to provide two things:
1) Higher surface area for speedy evaporation of the solvent (i.e. water or EtOH)
2) Prevent “bumping” (what you refer to as “boiling over”).
The best way to prevent a bump is to first spin your flask, then pull vacuum (you might have to apply more and more vacuum in stages if you have a lot of solids in your flask), and DO NOT place it into the hot water bath until the flask is cool to the touch or has ice crystals forming on the exterior. After it has cooled, lower the spinning flask (still under vacuum) into the bath.
PV=nRT
Dave, Sounds like things have really moved fast since Manhattan in May! Excellent news. Maybe the Buchi guys in the UK can add more fuel to your fire… Look forward to seeing you again soon in London.
BTW, on the subject of cold-finger probes, for Oxley we use Polycold to chill at -100C, LN proving too cold for our distillation process.
Best, M
Hey, Dave.
Great posting. Very interesting thread. You have given me a lot of new ideas for independent, out-of -the-box ideas for extractors/evaporators.
I think Sigma-Aldrich makes an adapter that might work for your residue removal: part #Z153141. It could be inserted between the vapor tube and your evaporation vessel, and draw the residue out of the flask/beaker/jar, whatever. But you would have to stop rotation and attach a hose to the side arm drain vent to use it. And I have not looked closely enough to see if it would be compatible with a continuous feed set up.
They also make an adapter for standard jars, part #Z723169, that is very reasonably priced when compared to the $500-750 that Buchi charges for their beaker-flasks. I believe they can be used with a wide range of shapes and sizes of jars. I am just about to go and check them out. How nice to be able to finish evaporating a concoction right in its storage/sales container!
Thanks for all of your great ideas. Very inspired and inspiring. Hope you keep it up. Although I imagine you are like me, in that with this kind of creative/inventive endeavor you cannot be deterred…
With my best from sunny Santa Fe, NM
Steven
Dave,
Greetings from Munich ! We use another rotovap model at http://www.derraum.com.au but I’m thinking of buying a Buchi for a new project in Munich. What chiller would you recommend? I noticed their big chiller only goes to -10. The system we have in Melbourne ( although it’s been worked on) can achieve much lower temps but I don’t have those lab rats here at my disposal in Germany so I need to buy a commercial system. I normally cook at around 35-65 but to be honest there is a lot of guess work involved and lots of wasted stock. The results are great but there is probably a better way, what reference point do you use when determining temp and pressure for herbs and other flavouring agents? Anyway continue to love reading about your work and look forward to running into you again on the speaking circuit. All the best Matt
Howdy! I use the Polyscience chiller, which will do -20 or better. The one thing I don’t like about it is it can’t take too much back pressure on the system so you can’t chill two items in series with it. I run alcohol as low as I can. Usually, by the time you are distilling water, the amount of energy required for condensing the steam onto the coils is so large that you won’t get freezing problems unless your distillation stalls. If you start getting freezing (wich isn’t the end of the world you can raise the temp a bit and the ice will disappear real fast.