Sciencemadness Discussion Board

Drying solvents with Al.

Furch - 2-10-2007 at 13:40

An idea occured when I read the current Al/Hg reduction thread.

Since pure aluminium reacts readily with water to form Al(OH)3 and H2, would it not be possible to use amalgamated aluminium to remove a whole lot (or all) of the water from chemically "inert" solvents prior to moisture sensitive reactions? Solvents such as THF, Et2O... But also to remove water from alcohols, such as ethanol.

Al + EtOH -/-> Al(OEt)3 + H2
Al(OEt)3 + H2O -/-> Al(OH)3 + EtOH

This method is used to dry ethanol, only magnesium is employed instead (Vogel's). Why not aluminium? It's a hell of a lot cheaper than both magnesium turnings and sodium wire.

Of course, amalgamating the aluminium with mercury salts is potentially unpleasant, but still.

What say ye?

Nicodem - 2-10-2007 at 13:50

It mostly depends on how effective it is in drying THF or other inert solvents. I doubt it can approach the effectiveness of a sodium wire and it certainly pales in comparison to the benzophenone ketyl radical drying method generally used. The dissolution of aluminium, in my experience, stops as soon as it gets covered with too much alumina slurry. You would need a terribly effective mechanical stirring in order to remove the hydrated alumina from the surface of aluminum amalgam and still it would take days for it to react with most water. On the other hand you might be able to dry it enough for a Grignard or some other such less demanding reaction. It is in the practical aspects and the apparatus required that I see problems.

Edit: I only now realize you are talking about drying ethanol. I don't think that would work well. Al(OH)3 might be soluble in ethanol by forming some kind of equilibriums among possible complexes. For example, you could end up with partially hydrolyzed species like Al(OEt)2OH which are likely to get solvated by ethanol instead of precipitating hydrated alumina. This would cause the same problem due to which sodium can not be used for drying ethanol (see Vogel for the explanation).

[Edited on 2/10/2007 by Nicodem]

Furch - 2-10-2007 at 14:04

I hear you Nicodem. Slurry'ing would be a real problem...

I guess drying the azeotropic mixture of THF-water with only Al would be pretty useless... But as a final touch it could work. I.e. after drying with KOH, which removes quite a bit of water in my experience. Then only a smaller amount of Al needs to be used, and hence intense magnetic stirring could work.

After all, one mole of Al removes three water - for drying one litre of KOH dried THF, I'd say it would be enough with a few grams of Al and a couple of days of reflux to remove the last accessable water.

I visit my lab about once every two weeks or so, so time is on my side.

I will give it a go someday ahead and report back! I'll test the THF water content with sodium metal when done.

Unfortunately I don't have any benzophenone lying around... :(

I just think it's a shame everytime I need to use sodium metal to dry my solvents, as I'd like to use it for other things.

What is your opinion on using activated magnesium as drying agent for ethers?

Thanks for the quick reply! Any further pointers are welcome, of course.

[Edited on 2-10-2007 by Furch]

Panache - 10-12-2007 at 21:24

i was fairly sure of reading in an older vogel (3rd i think) of drying 96% ethanol by distilling it over an al amalgam after first refluxing it over the amalgum. a decent reflux serves as an adequate mixer in this system (i have used this method). However i never did assay the ethanol i dried but it worked fine in the applications i used it for and wet ethanol wouldn't have, so this was enough evidence for me.
The setup is predictable, distillation vessel->column->stil head->condensor->collection vessel->drying tube. Replace the collection vessel after some 20ml is collected as this will have some water from the residual on the setup, take the replacement one straight from the oven at 150C and fit it while hot.

voila

come to think of it using al powder, added slowly would work much more quickly than foil or fine wire but requires more care, but i'm not game to try, besides i have a furnace now and quicklime is the way for me.

anyway thats what i know

not_important - 10-12-2007 at 21:49

Remember that aluminium reacts with alcohols to form Al(OR)3. The low solubility of Al(OH)3 will drive the equilibrium 3ROH + Al(OH)3 <=> 3 H2O + Al(OR)3 to the left, but once most of the water has teacted the alkoxide formed will quickly spread throughout the liquid, quickly zapping remaining water. And excess alkoxide is not going to distill over as it is rather non-volatile.

kmno4 - 11-12-2007 at 03:50

Al alkoxides are rather volatile.
It should be clearly described what amount of water we want to remove from solvent (and what solvent). Procedures are different in the case of <1% and 5% or 10%. Drying ~95% ethanol with Al(Hg) is really big pain in the ass :P. You will get a kind of cake with a lot of C2H5OH strongly adsorbed on Al2O3 hydrates. Even destillation this mixture is painful because of overheating and slow desorbing of etanol from Al2O3xH2O. Concentration of H2O is relatively high in 95% EtOH (count it in mol/mol)
If you want to remove small (<1%) amounts of H2O from ethanol, you can use Al(Hg) with no problems. For ethers, Na is very good.
The only cheap procedure ( I know) of drying ~90% sovents (rather solutions) is using CaC2. I wrote a post abot it and do not want to write the same here. My denurated EtOH dried with carbide react easily with Mg (+I2), so it is dry enough. The fact that it has not special pleasant odour, has no matter for me.
Large amounts of water can be removed from THF or dioxane in the same way. It is also a test for water content (brown-red CaC2 -> white Ca(OH)2 ).
BTW:
Using benzophenone in EtOH solutions is (of cource) useless.

not_important - 11-12-2007 at 04:14

Al(OEt)3 - BP 320 C, 210 C @ 10 mmHg, 180 C @ 4 mmHg
mp 154 to 159 C

The alkyls are more volatile, Al(Et)3 has a bp of 186 C

MgSO4 will work to rough dry EtOH, Al is indeed a finishing agent. Don't distill from the main reaction mix, filter and add a bit of fresh Al-Hg, reflux, distill - not to dryness.

Calcium carbide does work well, too. It has become difficult to get in some places.

chemrox - 11-12-2007 at 15:17

Quote:
Originally posted by Nicodem
...it certainly pales in comparison to the benzophenone ketyl radical drying method generally used.


How is this done exactly? I need to use this soon..ph-Li involved..

contrived - 12-12-2007 at 16:46

I can't find it either; I tried Perrin & Armego and Vogel's 3rd.

not_important - 12-12-2007 at 17:04

http://www.chemistry.msu.edu/Safety/safety_sop_01.asp

YT2095 - 14-12-2007 at 06:23

Quote:
Originally posted by kmno4

Besides - have you ever tried drying EtOH with MgSO4 ? Iam sure you have not, because it just does not work :P


I have actually tried this myself and I have to agree, the results were Very poor :(

I`m not going to say it Doesn`t work, it just didn`t work for ME.

kmno4 - 14-12-2007 at 08:30

So let us say that we both (and propably thousands others) have done this wrongly....
Let us wait untill someone will succeeed (~95% -> ~99% with MgSO4). In the meantime I will watch my growing beard.... becoming white....

;)

kmno4 - 14-12-2007 at 08:35

Quote:
Originally posted by not_important
Al(OEt)3 - BP 320 C, 210 C @ 10 mmHg, 180 C @ 4 mmHg
mp 154 to 159 C

The alkyls are more volatile, Al(Et)3 has a bp of 186 C

MgSO4 will work to rough dry EtOH, Al is indeed a finishing agent. Don't distill from the main reaction mix, filter and add a bit of fresh Al-Hg, reflux, distill - not to dryness.

Calcium carbide does work well, too. It has become difficult to get in some places.

BP 320 C means what it means. Compare it to NaCl, CaC2, NaOH, K2SO4... or any ionic/polymeric compound. I did not say that such alkoxides boil at 50 C or so. Al alkoxides are rather volatile compounds and can be easyli purificated by (vacuum) distillation.
Besides - have you ever tried drying EtOH with MgSO4 ? Iam sure you have not, because it just does not work :P

BTW:
Interesting article (free) about benzohpenone salts of Na can be found here:
http://www.rsc.org/publishing/journals/DT/article.asp?doi=b5...
It it worth to dowload "Electronic supplementary information" in CIF format and open it in Mercury (free) program, to see nice 3D (+a few options) structures of solvated Na salts.

**** this is my second post in this thread (a little mistake with deleting..)

solvent stills as related to thread

chemrox - 19-12-2007 at 19:06

"To keep stills dry and oxygen free, they are connected to an inert gas source, and fitted with a bubbler to vent excess pressure."

Are we ageed that a baloon works as well and has the added advantage of maintaining positive pressure on the system?

and:

do I have to put benxophenone and Na in my ultrasound cleaner to make the ketyl? or do I reflux the two? how long? until the color is deep blue? ie, how do I make the stuff? nevr done it .. used Na in the past or CaO for alcohol...

[Edited on 19-12-2007 by chemrox]

cbfull - 20-12-2007 at 11:04

What about passing the vapor phase over a column of silica gel, or even drierite? Silica gel can have bad reactions to liquids, but a vapor should be fine.

Has this been done before? Insight?

Furch - 29-12-2007 at 02:56

The point was to use Al because it's dirt cheap and ridiculously available. Also, if it works, it will probably remove nearly _all_ of the water. MgSO4 is a really crappy drying agent when trying to make something close to anhydrous.

I will give the Al method a try sometime, and report back to you my findings.


- Furch

dangator - 24-1-2008 at 23:36

The use of alumina columns for drying solvents is widespread, the original credit goes to Grubbs for phasing out solvent stills. Maybe three phase drying using something stupid (MgSO4), distillation, followed by running though an alumina column would work wonders. The alumina can likely be reactivated with excessive heating. Even easier is just using molecular sieves of the appropriate size--they're incredible dehydrating agents that receive less credit than they deserve. Of course you'll want to do one of these to quantitate what's happening:

http://en.wikipedia.org/wiki/Karl_Fischer_titration

Panache - 25-2-2008 at 03:56

buy a air-conditioning drying canister from a refrigeration supplies and run your solvent through it. They are filled with mol sieves and if you choose the right ones (read the catalogue) they are idiot-proof, well almost.
They are designed to remove water from refrigerants and are usually fitted in-line just after the compressor so the refrigerant is a liquid as it passes through them.
One the size of a can of soft drink generally mops about 20mL of water. ensure you don't break the wax seal until you use it though as it will grab water from the air. I haven't tried opening one before to regenerate it.