Sciencemadness Discussion Board

How can I prevent excess solvent loss?

Commodore_Commonwealth - 3-3-2013 at 21:55

Ive been using an aspirator pump to trap pull a reasonable (~45mtorr) vacuum for a distillation that I routinely run. I am using a standard distillation setup with as large a condenser as I could find (maybe 18 inch across?). Im pumping icewater through the condenser from the same res as the aspirator pump and have a receiving flask suspended in salty icewater. The issue is that I routinely loose 20% or so of my starting solvent, which adds considerable cost to common reactions. Using dryice instead of salty icewater did noticeably decrease my losses, though the cost is still pretty considerable.

I have a glass gas wash bottle, is it safe to use this as a second cold trap? I've never seen anyone do this so I am a bit suspicious.

radagast - 3-3-2013 at 22:45

What solvent are you removing?

Probably wouldn't hurt to give a gas washing bottle a try, although I'm not sure that it's really designed for low-pressure applications. I do have a dedicated refrigerated -60C cold trap which uses a piece of glassware similar to a gas washing bottle, except substantially bigger, and with thicker-walled glass.

Have you tried using a secondary condensor?

[Edited on 4-3-2013 by radagast]

GammaFunction - 4-3-2013 at 02:07


Is it necessary to use the aspirator to maintain vacuum? Or can the glass be sufficiently tightly sealed so that it remains at vacuum for a finite length of time? If the latter, could one close a valve to the aspirator for most of the time?

I would think that the only sources of vacuum loss would be external leaks and dissolved or generated gases. Greased glass is supposedly resistant to the former, and dissolved gases would come out of solution early on.

Dr.Bob - 4-3-2013 at 08:23

It is nearly impossible to run a vacuum distillation without a constant vacuum source. No glassware is leak proof, and it is almost impossible to control the vacuum in a sealed system. Just take my word for it, it does not work.

Using a simple extra trap will help some, but all distillations lose some material, the best way to lower the loses are 1) good condenser and cooling, 2) extra trap, 3) larger scale, so less material is lost in glassware, still bottoms, etc, 4) good design of apparatus. Most distillations will lose 10% with best design; much higher for poorly functioning systems. Depending on scale, it may be hard to improve your by a lot, without spending more on traps, dry ice, etc.

GammaFunction - 4-3-2013 at 12:10

Quote: Originally posted by Dr.Bob  
It is nearly impossible to run a vacuum distillation without a constant vacuum source. No glassware is leak proof, and it is almost impossible to control the vacuum in a sealed system. ....


To stubbornly play devil's advocate, the vacuum is 45mbar [see note below], which isn't terribly low. Well greased round glass joints are claimed to be able to hold a vacuum for years. My Teflon joint sleeves are advertised to being good to 1 mbar or so.

So it seems like the leakage rate could be brought pretty low. Therefore, could an intermittent vacuum work with the pressure changing cyclically in some range (say, 40 to 50 mbar instead of a constant 45? If the vacuum is on 10% of the time, would not losses be 10% of current losses?

I imagine that what an aspirator does at its operating limit is that it 'trades' outgassed air and water vapor for solvent molecules. If this inflow of air from this trading process dominates the leakage from the joints, would not intermittent pumping give a big advantage?

(Note: The original post says the vacuum was 45 mtorr using an aspirator. This seems incorrect, as an aspirator can do only 30 torr or so. Maybe 45 mbar was meant?)

Dr.Bob - 4-3-2013 at 13:47

I took it to mean 45 torr. You might be able to do an intermittent vacuum application, just like a triac dimmer switches the power on and off quickly. But in my years of work, I have never heard of a vacuum distillation that was able to be usefully completed without being connected to a vacuum during the entire experiment, not to say that you cannot control it via pulsing or intermittent use. Most of my experience with that comes from two different ends of the spectrum, both higher vacuum small distillations, as well as rotovap type distillations, with a lower vacuum, but higher throughput.

I have had no luck doing either without a good vacuum source that allows me to create a constant vacuum during the run, otherwise, if your vacuum is not connected, air will slowly leak in and kill the vacuum and the temperature will have to be continually raised during the distillation, which will not allow you to obtain decent purity of the distillate.

And most rotovaps leak like a sieve, so they are much worse. But for normal use, that slight flow of air helps move the vapors from the flask to the condenser, so it does not hurt much unless you are trying to remove DMF or the like. As long as the vacuum pulls a little flow, that is all that is needed.

Glassware joints leak some, and Teflon joint sleeves leak worse than grease, you will not be able to HOLD a 1mbar vacuum without constant application of the vacuum. You should be able to create a 1mbar vacuum using them along with a vacuum pump, but not hold it for long without. At least not for a normal distillation apparatus with several joints. Remember, nature abhors a vacuum.

GammaFunction - 4-3-2013 at 16:13

Quote: Originally posted by Dr.Bob  

Glassware joints leak some, and Teflon joint sleeves leak worse than grease, you will not be able to HOLD a 1mbar vacuum without constant application of the vacuum. You should be able to create a 1mbar vacuum using them along with a vacuum pump, but not hold it for long without. At least not for a normal distillation apparatus with several joints. Remember, nature abhors a vacuum.


I agree. My argument is if you can make a 1 mbar vacuum and hold it as it rises to 2 mbar over, say, a minute, then you should be able to hold a 40 mbar vacuum as it rises to 50 mbar for 10 minutes, allowing intermittent pumping (which would admittedly throw off fractionation).




ziqquratu - 5-3-2013 at 01:59

I often run our rotavaps with "one-shot" vacuum. They seal well enough that if I drop the pressure to, say, 100 mbar below that required for efficient evaporation of the solvent, then close off the pump, the pressure will remain fairly constant, since the rate of evaporation is balanced by the rate of condensation. It works really well if I have a slightly volatile compound, so I don't lose it to the rotavap.

I've had success with a range of solvents, most commonly ether and dichloromethane (i.e. low boiling), but also ethyl acetate and similar. I've never tried toluene, but I have no reason to think it wouldn't work if I increased the bath temperature a little. That's also without applying ANY grease to the rotavap - meaning the joints between flask and splash bulb; splash bulb and rotor; and condenser and collecting flask.

I think it comes down to how well your apparatus is sealed, how much solvent you're taking off, and what that solvent is. But, since you're claiming to lose ~20 % with a 45 torr (like Dr.Bob, I assume the "m" was a typo!) vacuum, it must be reasonably volatile, so it might work for you.

Organikum - 5-3-2013 at 02:37

How can something be lost to the pump when the system is sealed?
Logic says there is a minisculous amount to none flow of matter when a constant vacuum has been reached. Only if the seals leak something moves. Or not?
So if nothing leaks you dont have to seal off the vcuum pump for nothing moves there (not in any amount worth mention). If your apparatus is not sealed as perfectly matter will move, but then you cannot disconnect the pump anyways.

After my experience when vacuum is diminished by just alittle distillation breaks down and there is just a hot liquid sitting there with about nothing coming over. Maybe a Büchi rotovap or other such equipment can do but a regular setup can not in most cases. Except for some very docile compounds excessive heat causes excessive boiling, bumping or other unnecessary stress on the material and apparatus. For it si mostly a rather tight window to work in, but maybe I am just doing things wrong.

/ORG

Nicodem - 5-3-2013 at 08:47

Try to use a bleed valve to increase the distillation pressure and thus reduce losses.

ziqquratu - 5-3-2013 at 19:24

Organikum, I assume your post was directed at me?

Trust me, it works. I can post a video (one day...) if need be! Don't get me wrong, I often need to apply the pump a couple of times to get all the solvent to go, but the flow rate is really very good. I think the key is to ensure that the pressure you get it to is below that required for evaporation at the bath temperature, and that you ensure it starts distilling before you shut off the pump. If you don't do the latter, the pressure rises as vapour fills the rotavap, which then prevents further evaporation, while the former provides a buffer against leaks.

My assumption is that it works through an internal pressure gradient, caused by condensation of the solvent nearest the condenser creating a local region of low pressure, drawing more vapour toward the condenser, and so on.

You are right, though, I'm not sure how well it would work in a regular distillation apparatus. I should probably have made that clear!

GammaFunction - 6-3-2013 at 02:46

Quote: Originally posted by ziqquratu  

My assumption is that it works through an internal pressure gradient, caused by condensation of the solvent nearest the condenser creating a local region of low pressure, drawing more vapour toward the condenser, and so on.

You are right, though, I'm not sure how well it would work in a regular distillation apparatus. I should probably have made that clear!


I don't think there is anything strange about distillation working in a well-sealed apparatus under vacuum. It's like distillation working at atmospheric pressure - everything is at 1 bar, with no vacuum pump. Why not less than 1 bar?

The vapor pressure of hot fluid is higher than that of a cold fluid, so there is a net flow of vapor from hot to cold.

Commodore_Commonwealth - 6-3-2013 at 13:29

Thanks for all the suggestions guys, I really appreciate your thoughts. After looking at my options, Im going to run my distillation at a reduced vacuum to slow down the sublimation and hopefully trap more condensate. I suspect Im pulling it a bit too hard. Id rather not have to pulse the pump, though after talking it over with others it sounds like its an approach that is viable.

Organikum - 6-3-2013 at 14:40

Quote: Originally posted by GammaFunction  
Quote: Originally posted by ziqquratu  

My assumption is that it works through an internal pressure gradient, caused by condensation of the solvent nearest the condenser creating a local region of low pressure, drawing more vapour toward the condenser, and so on.

You are right, though, I'm not sure how well it would work in a regular distillation apparatus. I should probably have made that clear!


I don't think there is anything strange about distillation working in a well-sealed apparatus under vacuum. It's like distillation working at atmospheric pressure - everything is at 1 bar, with no vacuum pump. Why not less than 1 bar?

The vapor pressure of hot fluid is higher than that of a cold fluid, so there is a net flow of vapor from hot to cold.


Of course you can disconnect the pump from a decent setup for quite some time, otherwise an Anschütz-Thiele Vorstoß for example would not work and I could not use the toilet during distillations (as openeing another faucet in the vicinity lets the aspiartor starve and vacuum suffers.
The question is - does shutting off the pump prevent loss of solvent?
It seems to be a question of design and not of physics, in the rotovap design it works, another design may have no need for this at all.
Above is stated that flow is from hot to cold, if this is true nothing should get lost to the pump for the condensator is the coldest region around (if a trap is used then latest there should be anything collectable collected.
I support the idea that the design of the setup is the main factor for solvent loss or no solvent loss. For actually I found also solvents vanishing as by magic in big amounts. I also saw it go down the drain. The vacuum connectors where the tube connects to the glass, thats the point it happens. Many of these parts are actually generating a tin film of destillate which of course is sucked away by the slightest movement of air. Better glassware has longer and not jst straight but angled tubes for the condensate trying to avoid the problem, but thats not easy and is a crutch nothing more.
I got the Anschütz-Thiele not so ong ago and I wont miss it now, never again. Get something like this, and the problem should be solved. That was one of the first things I mentioned - virtually no losses to the pump anymore!
And not to forget it is a great plaything with all the valves and it is astounding how many WRONG combinations exist and how FAST one is to find every single one.

/ORG

bahamuth - 6-3-2013 at 15:06

Quote: Originally posted by Dr.Bob  
But in my years of work, I have never heard of a vacuum distillation that was able to be usefully completed without being connected to a vacuum during the entire experiment, not to say that you cannot control it via pulsing or intermittent use.


Read somewhere that one guy had good experiences with attaching an empty flask after the condenser in such a way that it would not gather any condensate, add an amount of KOH, purged the setup with CO2, and evacuated the whole setup as much as he could before closing it completely and letting the KOH eat up the rest of the CO2, achieving down in the tens of thousands of a bar even without a multistage vacuum pump. And he did not evacuate it any more during the run as I understood it.

As the theory goes, if you evacuate the system until it has reached thermal equilibrium (e.g. steady distillation) one does not need any "more" vacuum. That being said I have never had a completely airtight setup, even with really really viscous joint grease. Perfluor wax would be better at high temps and vacuums and I want some of that if anyone got some to sell cheap..

PS. Regular rotavapor losses amount to around 20% for low boiling solvents, and this can be mended somewhat (down to around 5% IIRC) by pulling the vacuum from an (vacuum)adapter between the receiving flask and the condenser on a standard rotavapor setup, not on the "top side" as most originally do. Can't find the reference for that though, but found it on some online lab technique journal.

Try to evacuate the setup once it has reached equilibrium and close it while monitoring pressure, I guess you will do fine as long as you don't run the for extended periods at a time.