Sciencemadness Discussion Board

Anhydrous Diethyl Ether via NaOH?

teragon - 4-6-2009 at 21:30

Now, I've always been under the impression that short of sodium wire/metal (or maybe POCl5)- nothing else will really suffice for making anhydrous diethyl ether (for use in Grignard reactions, etc).

I'd rather not purchase expensive sodium and even then- I would hate to waste such a useful compound for dehydration.


I recall someone on here mentioning that dry NaOH could be used to obtain satisfactorily anhydrous, Grig-useable ether.

I think it was Organikum, but i don't want to point any fingers. So if anyone could shed any light on this, I would appreciate it.


Of course, I would prefer references/papers on the subject, rather than secondary anecdotal evidence- but anything is helpful, welcomed. Thanks in advance. Anhydrous ether is just too damn expensive.

entropy51 - 5-6-2009 at 05:29

Organikum indeed mentioned using NaOH to dry ether for Grignards in this thread:

http://www.sciencemadness.org/talk/viewthread.php?action=pri...

He cited no reference that I can see, but there is a paper from J Chem Ed posted on the forum describing NaOH as a dessicant for isopropanol, and another paper from J Chem Ed describing MgSO4 as a dessicant for ether, and yet another for MgSO4 to prepare anhydrous HCl in ether.

A search should turn up those papers. Use the search button at the top of the page, or better, use Google with www.sciencemadness.org/talk.

Eclectic - 5-6-2009 at 07:11

3A or 4A molecular sieves should work also....but if done in an ultrasonic cleaner, even wet ether works.

Hydragyrum - 5-6-2009 at 07:26

NaOH may well work - you then make an excess of Grignard reagent and some of it is taken up drying the solvent completely (the solvent will need to be of a certain level of dryness for this to work, and I'm not 100% sure NaOH can achieve this - best just to try it and see!).

teragon - 5-6-2009 at 07:28

Entropy- thanks for that thread. i hadn't seen that one before, but recall organikum mentioning it in another thread.


Quote:
Organikum said:
NaOH or better KOH for drying ether for a Grignard.And enough time. Period. Why? Some guys tried this out long time ago and published the results. Ok?


I would love to see this paper- if Organikum/anyone knows/remembers the paper reference/citation, please share.


Figured anhydrous KOH would work well too, then. Good.


Eclectic- Most of us don't have ultrasonic cleaners hanging around, plus I'd prefer to do the reaction anhydrous for best yields.

MolSieves is a good idea, I'd imagine they should work- but then again, I have no evidence saying they do or don't. I don't have any, but if I knew they worked then it'd be a sound investment considering they can be reused.

Saber - 5-6-2009 at 07:31

Firstly Sodium is not a compound.
Secondly anhydrous Sodium hydroxide does suffice for drying ether, but it must be Anhydrous, which is fairly difficult to buy, but can be made by heating well beyond melting point and holding it there for around 30 mins in a CO2 /H2O free atmosphear (preferably over Argon)

DJF90 - 5-6-2009 at 07:33

Mol sieves definately work. For anhydrous ether, you want 4A. For anhydrous MeOH or EtOH, use 3A.

teragon - 5-6-2009 at 07:36

Semantics. I should have used the word element, more appropriate. But compound has become a commonplace english substitute for any "material". Not just "compounded" things.

I hate language.


Anyways- any papers/evidence? Have you done this yourself or are you just repeating what you've heard?

teragon - 5-6-2009 at 07:39

Quote: Originally posted by DJF90  
Mol sieves definately work. For anhydrous ether, you want 4A. For anhydrous MeOH or EtOH, use 3A.



Thank you :D

Hydragyrum - 5-6-2009 at 07:40

Another thing: NaOH may well be fine for drying alcohols but I'm no longer sure if it is good for ethers - it is conceivable it may react with them:

CH3CH2-O-CH2CH3 + NaOH --> CH3CH2-OH + CH3CH2-ONa

Not sure how likely this is, but I do remember using NaOH for drying amines. I have only ever used Na, K, or mol. sieves for drying ethers.

I doubt MgSO4 can get ether dry enough for a Grignard reaction to start though.

PS. A good way to dry alcohols which are liquids* is with Mg and iodine, which generates first Mg(OR)2 and then reacts with the water to give Mg(OH)2, which precipitates. You then distil the alcohol from this.

* Some alcohols are solids.

bfesser - 5-6-2009 at 08:18

Quote: Originally posted by teragon  
Semantics. I should have used the word element, more appropriate. But compound has become a commonplace english substitute for any "material". Not just "compounded" things.

I hate language.


Anyways- any papers/evidence? Have you done this yourself or are you just repeating what you've heard?


If you're on a chemistry forum asking about how to prepare anhydrous diethyl ether, yet you don't know the difference between a compound and an element...

Anyways:
http://en.wikipedia.org/wiki/Chemical_compound

entropy51 - 5-6-2009 at 08:42

There are threads on the use of molecular sieves, lots of ether threads. As they say around here, Use the Fantastic Search Engine.

If I did it right, this link should bring up the paper on the use of MgSO4 for drying ether for Grignards: http://www.sciencemadness.org/talk/files.php?pid=128825&...

In addition to ultrasonicating, there are some other tricks to start a Grignard when the ether is not absolutely, perfectly dry. One is to heat the magnesium in the empty reaction flask with a crystal of iodine to activate the magnesium surface. Another that usually works is to start a micro-Grignard in a test tube with 1 mL of methyl iodide in a little ether and some Mg and when that starts (due to the high concentration of MeI), dump it into the main Grignard. That scavenges traces of water and more often than not it starts. I have used this trick in humid summer weather when just pouring the ether causes it absorb atmospheric moisture.

Hopefully Organikum will have something more to say about using NaOH.

teragon - 5-6-2009 at 09:13

entropy-
I used the f-ing search engine for NaOH/ether, which is what I wanted information on. ;)

MolSieves is a nice side note, just a bit expensive.

Thanks for that MgSO4 paper/abstract- it recommends MgSO4 for the "redrying" of anhydrous ether. Meaning for drums of anhydrous ether that have absorbed atmospheric humidity....not wet/used ether.

But i guess if you thoroughly dried it with CaCl2 first, then MgSO4- perhaps it's strong enough? Only one way to find out I guess...The differences they noted between sodium reactivity with MgSO4-dried ether vs. CaCl2-dried ether seems promising.


bfesser- Thanks, but I know what a compound is. I was using the word colloquially- I forgot how pedantic some people are when it comes to definitions/communication. My bad.

entropy51 - 5-6-2009 at 10:23

Wait and see what Organikum has to say about the NaOH.

Sieves are pricey, but they can be dried and re-used. In another thread Magpie said:

"I make my own ether using a 200 mm West condenser (like a Liebig) and ice water cooling for cooling water. I don't make it very often and only in small quantities so I don't have to store it.

I dried it with 3A molecular sieves and it worked perfectly for a Grignard."

And you noticed that the paper says that ether dried with MgSO4 hardly reacts with Na at all - which would suggest that further drying on Na is superfluous!

You could make a small trial run, drying first with CaCl2 and then MgSO4. It does work if the halide is not too unreactive and if need be can be primed with MeMgI as I described. Or make a trial run drying with NaOH. The best way to find out if something will work is to try it on a small scale. Even if it worked for someone else it may not work for you, and vice versa.

panziandi - 5-6-2009 at 11:05

I personally have used NaOH 20% solution to remove peroxides in ether, the washings quickly reduce peroxide content to below detectable levels with KI method. As to drying with solid NaOH for a Grignard I wouldn't. My method would be:

1-(NaOH wash to rmeove any peroxides)
2-Preliminary drying with CaCl2 or MgSO4
3-Drying with Na (and distillation from Na)

Can use NaHg NaPb etc also (former being liquid, later being solid and brittle)

Nicodem - 5-6-2009 at 12:07

Quote: Originally posted by teragon  

MolSieves is a good idea, I'd imagine they should work- but then again, I have no evidence saying they do or don't. I don't have any, but if I knew they worked then it'd be a sound investment considering they can be reused.

I assume you know that you can only get as precise an answer as precisely you set the question. Asking questions in such a general sense of "will NaOH dried ether work in making the Grignard reagent" is not particularly precise enough for a definitive answer.
Some Grignard reagents can be made even with undried ether and some luck, while for many even MgSO4 dried ether will work with or without iodine surface etching. Molecular sieves will do fine for most substrates, but then there are some substrates that will not form an organomagnesium reagent even if you use benzophenone ketyl dried ether and ultrasound to break the surface of Mg. It is also important which type of magnesium you use, or which of the numerous Grignard tricks you apply (these are some type of lab folklore, ranging from iodine tricks to Fe salts addition... even prayers can used).

teragon - 6-6-2009 at 23:00

Good point, ok then. I will be more specific. Since I have your attention...mind sharing some knowledge? :) Something common, like BenzylMagChloride.

Do you reckon MgSO4 or NaOH dried ether would suffice for benzylmagnesiumchloride preparation? Iodine surface etching is employed. Pure magnesium saw chips (washed with ether) are used.

Good results with UNIVAR AR-grade, anhydrous ether have been observed, even when mixed with dry xylene (ether:xylene - 3:1). Also...How far do you think one could stretch the ether:xylene mixture when preparing said reagent?

Thanks

Nicodem - 7-6-2009 at 08:06

The preparation of a Grignard reagent with benzyl chloride is commonly easy and is not particularly sensitive to a reasonable amount of moisture. Molecular sieves should be more than enough. NaOH, KOH or MgSO4 probably, but why don't you just try?
Do you know how to start and run a Grignard? You must heat dry the glassware and use Grignard grade magnesium turnings or powder. Adding a crystal of iodine is always a good idea, also in order to visually give you some information on how the surface etching and Grignard initiation is proceeding. Molecular sieves can also be used for in situ drying if you want. I do not know on what scale you will be doing this, but if anything more than a dozen mmol then make sure you do not add too much BnCl at the beginning when waiting for the reaction with Mg to start (or else you will be witnessing how a Grignard runaway reaction looks like).
I can not answer your question about how much xylene can you admixture with impunity, because you did not say what reaction on which substrate will you be using your BnMgCl solution for.

entropy51 - 7-6-2009 at 08:30

I wasn't aware of using xylene to replace part of the ether in a Grignard, but another trick I've used to replace part of the ether is worth mentioning. Instead of adding the reactant to the Grignard in ether, it can be added in dichloromethane instead. This works for aldehydes and ketones and surpisingly enough, actually increases the yield.

I wish I could take credit for this trick, but it's in the literature. Aldrich once pointed it out in an ad.

I second Nicodem's caution about not running in the BnCl until the reaction has unambiguously started. There are few things worse than a bath of ether and BnCl.

[Edited on 7-6-2009 by entropy51]

bfesser - 7-6-2009 at 14:16

Quote: Originally posted by teragon  

bfesser- Thanks, but I know what a compound is. I was using the word colloquially- I forgot how pedantic some people are when it comes to definitions/communication. My bad.


That's not a colloquialism, that's just ignorance.

teragon - 7-6-2009 at 23:29

Nicodem-
The reason I was asking before trying- is because my meager pockets don't want to shell out funds for more ether or invest time/reagents in synthesizing it- until I know that it can be dried thoroughly for said reaction.

Now that you've indicated that it is likely feasible, I will bother to invest time/reagents/etc.

And yes, I know how to start and run a Grig. The reactions I was reporting above were performed by myself. Yes, Grig-runaways are always best avoided ;) Not to mention, over addition of BnCl will lead to coupling of the Grig with BnCl.

As for the rxn substrates- they are all completely soluble in xylenes. My question was about preparation of the grig, not the eventual reaction. I know that once the reagent is prepared, most of the ether can be replaced with another solvent (in some cases atleast). I recall a patent in which 1.0M BnMgCl in ether was combined with kerosene during the evaporation of the ether- for addition and refluxing of a ketimine to/with BnMgCl at higher temps.

I am just curious to what extent the ether molecules are necessary for proper Grignard complex formation- I know they play a vital role in coordinating the Mg molecules, but I'm not exactly sure how much is required for formation of each "R2Mg-MgX2" grig complex.



Entropy-
The same applies for xylene and various other solvents. Not only can BnMgCl be prepared in ether:xylene - 3:1 ratio...but I have then added xylene soluble reactants (in xylene) to such a prepared grig-reagent with no affect on yields (compared to a "pure ether" grig prep AND rxn). It's a pretty common technique, I believe.


bfesser-
If you want to argue about frivolous things like definitions and proper word-usage, please post elsewhere. I admitted my mistake- you obviously have insecurity issues.

Nicodem - 8-6-2009 at 00:41

Quote: Originally posted by teragon  
Nicodem-
And yes, I know how to start and run a Grig. The reactions I was reporting above were performed by myself. Yes, Grig-runaways are always best avoided ;) Not to mention, over addition of BnCl will lead to coupling of the Grig with BnCl.

Coupling of BnCl with BnMgCl (the Wurtz reaction) is indeed the major problem when preparing benzylmagnesiums. Slow addition of BnCl helps to some degree, but avoiding reflux conditions works even better. I used plain BnCl only once to prepare its Grignard reagent and did not experience this problem, but probably because as soon as the magnesium started to react I immersed the reaction flask in an ice bath and continued the BnCl addition at a low enough rate to avoid temperature increase. With some other ring substituted benzyl chlorides (and particularly their benzyl bromide counterparts) I was not so lucky and there always formed from approximately 25-50% Wurtz product even with cooling. But I think you will not experience too much troubles with this side reaction if you will use plain BnCl. Always remember that on easy substrates such as alkyl or benzyl halides the Grignard is problematic only in getting it to start, once it starts it runs regardless of the temperature, so you can freely use an ice bath after its initiation.

Quote:
As for the rxn substrates- they are all completely soluble in xylenes. My question was about preparation of the grig, not the eventual reaction. I know that once the reagent is prepared, most of the ether can be replaced with another solvent (in some cases atleast).

The amount of ether needed for magnesium chelation is small and you can easily reduce the solvents ratio probably even to 1:1. The problem is not the solubility and ether chelation of BnMgCl, but rather in that the hydrocarbon solvents alter the kinetics of the subsequent reaction and cause the crashing out of the solution of the reaction product. If you will add your BnMgCl to a reactive carbonyl compound such as a ketone then altered kinetics are no problem, but you might experience complete clogging up of the reaction due to a voluminous precipitate of the magnesium alkoxyde/halide salts. This makes stirring impossible and may occlude the reactants from reaction (giving low and unreproducible yields).
Besides diethyl ether is just as cheap as xylenes.

teragon - 9-6-2009 at 10:13

Thanks very much for the insight and helpful tips. I figured the amount required for grignard reagent formation would be small. Understandable that the basically non-polar, hydrocarbon solvents would alter solubility and hence kinetics.

Extremely appreciated :D


Panache - 9-6-2009 at 19:57

I just bought an ultrasonic toothbrush, removed the bristle head, now i have a cordless rechargable ultrasonic stainless wand, works great.

Saber - 12-6-2009 at 02:14

Small ultrasonic baths can be bought off ebay for around £10.
They can be modified to make larger baths to fit 1l flasks.
Myn fits a 100ml rbf nicely, and is great for grignards!

zed - 21-6-2009 at 16:52

I've prepared Benzyl Magnesium Chloride on several occasions. Never had a problem.

Used the I2 method however. Put the Magnesium in my reaction vessel, then I fitted the vessel with a dropping funnel and a reflux condenser. Thereafter, I warmed the Magnesium in the flask, turned off the heat, dropped a crystal of Iodine down from the top of the condenser, dripped in a few drops of Benzyl Chloride mixed with Ether.....And Shazzam! The reaction always started immediately.

If you are not familiar with Benzyl Chloride, it is a lacrimator. And, while you CAN run a Benzyl Chloride Grignard reaction outside of a fume hood, you will probably find it very unpleasant.

As for drying wet Ether.....in the past, I found that Sodium wire alone, was inadequate for drying cold, wet, recycled Ether. At least, for Ether destined for LiAlH4 reductions. Drying by some other method, or pre-drying by some other method, is a good idea.

S.C. Wack - 18-7-2009 at 17:54

Well I didn't think that there were any more drying articles by Burfield and Smithers to find, but I found another one. Now this does not use NaOH, nor are they "Drying ethers over P2O5", but I mention this here because MgSO4, sieves, and CaCl2 are mentioned in this thread.

Here various things were tried drying wet (14.7 mg/g) ether, with a maximum drying time of 6 hours. MgSO4 (like everything else listed here, this was heated to 350C overnight, loading at 10% wt/vol with intermittent shaking) gave residual water content 1.8 mg/g.
CaCl2 powder, 0.24 mg/g. 0.39 mg/g in pellet form.

4A bead: 0.29 mg/g.
4A powder: 0.095 mg/g. Inexplicably, this is a higher number than the amount of water at 15 minutes (0.092), 30 minutes (0.076), and 1 hour (0.027, the lowest figure in the article, JCE 703, 1982)

5% loading = more water in all cases, especially with the sieve powder.

[Edited on 19-7-2009 by S.C. Wack]

peach - 28-8-2010 at 03:44

Curly arrow has an excellent blog entry on anhydrous solvents. Also have a read of the comments.