Sciencemadness Discussion Board

Preparation of “Absolute” Ethanol (<100 ppm water)

Magpie - 24-2-2018 at 12:56

Absolute ethanol (<100 ppm water) is often a requirement for a successful synthesis in organic chemistry. Every mole of water can result in a loss of a mole of product.

This procedure is based on that found in Vogel (ref 1). The use of 3A mole sieves is added to bring the water content down from ~1% to <0.01%, Ref 2).
‘
A. Reagents

325 ml of Everclear (190 proof ethanol, or “rectified spirit”)
107g of clean garden slaked lime (Ca(OH)2)
30g of 3A mole sieves
NaOH for glassware cleaning soln
silicone oil for oil bath (DOT 5 brake fluid works well)
CaCl2 desiccant for guard tube
HNO3 washed fiberglass wool

B. Apparatus

1 liter 2 or 3 neck RBF
condenser
vacuum adapter
3-way distillation adapter
500 ml alcohol storage bottle with ground glass fitting that fits the vacuum adapter
large ceramic crucible
muffle furnace
thermometer
hot plate
strong overhead stirrer
oil bath
insulating blanket and/or aluminum foil
guard tube
powder funnel

C. Procedure

Clean all glassware with an aqueous ethanol solution saturated with NaOH. Then dry for 2 hours at 130°C. This will produce sparkling glassware.
Here's the recipe for the cleaner: https://www.inrf.uci.edu/wordpress/wp-content/uploads/sop-we...


a. Dehydrating the Ca(OH)2

1. Weigh out 107g of slaked lime into a large ceramic crucible. Place crucible in a muffle furnace and set for 800°C.
2. After 3 hours remove the crucible and place in a desiccator to cool. Be very careful when removing lid from desiccator. Remove it slowly.

b. Reflux the Everclear

1. Set up an oil bath on a hotplate. Affix a 2- or 3-neck 1 liter RBF in the oil bath.
2. Add 325 mL of Everclear to the RBF. Add the CaO to the RBF using a powder funnel.
3. Install a reflux condenser in a side neck.
3. Install an overhead stirrer in the vertical neck of the RBF and begin stirring.
4. Set the oil bath at 100°C. Heat sufficiently to bring the slurry to reflux. The slurry will thicken considerably as the CaO takes on water.
5. After 6 hours at reflux turn off heat. Remove bath but leave the stirrer in the RBF. Leave condenser running until slurry is cool. Leave the guard tube in place in the RBF. Let sit overnight.

c. Activate the 3A Mole Sieves

Note: 3A mole sieves are rated at 22 wt% H2O absorption. Assuming 1% H20 left in the ethanol this would be 2.63g H2O. Assuming, conservatively, 15 wt% water absorption by the sieves, wt sieves required = 17.5g. Use 30g.

1. Set the muffle furnace to 300°C.
2. Place 30g of 3A mole sieves in a crucible, place in furnace, and heat for 1 hour..
3. Remove from furnace when T = 200°C and place in a desiccator to cool.
4. Pour the cooled mole sieves into the 500mL ethanol receiver and cap.

d. Distillation of the Ethanol

1. Setup for simple distillation using the silicone oil bath (100°C) and overhead stirrer.
2. Place HNO3 washed fiberglass wool in the horizontal leg of the distillation adapter per Vogel. This is to remove any entrained CaO.
3. Stir and distill the ethanol into the receiver with the mole sieves. Use an insulation jacket if needed.

D. Discussion

Using a strong, reliable overhead stirrer is critical for this procedure.

Although I have used this procedure at least 3 times I never did write down the yield of absolute ethanol.

E. References

1. A Textbook of Practical Organic Chemistry, 3rd ed, A.I.Vogel, 1959, page 166, “Dehydration of Rectified Spirit by Quicklime.”
2. JOC article: “Drying of Organic Solvents: Quantitative Evaluation of the Efficiency of Several Desiccants,” Williams and Lawton, University of Johannesburg, August 12, 2010.

Questions, comments, and suggestions are welcomed.



[Edited on 24-2-2018 by Magpie]

[Edited on 24-2-2018 by Magpie]

[Edited on 26-2-2018 by Magpie]

[Edited on 26-2-2018 by Magpie]

[Edited on 26-2-2018 by Magpie]

[Edited on 26-2-2018 by Magpie]

happyfooddance - 24-2-2018 at 13:07

I am confused. Are you using slaked lime or CaO?

Very detailed write-up, otherwise!

Magpie - 24-2-2018 at 13:10

What you buy at the garden center is slaked lime, Ca(OH)2. By heating it in the muffle furnace at 800°C it is converted to quicklime, CaO.

happyfooddance - 24-2-2018 at 13:16

My bad, I was reading too quickly through your post!

Why do you believe that a stirrer is critical? Dry boiling stones work for me.

[Edited on 2-24-2018 by happyfooddance]

Sulaiman - 24-2-2018 at 13:23

In C,a,1 where did the aqueous ethanol come from and what is its concentration ?
..........................................................................
In the writeup you set muffle furnace to 300 C but in a reply you mention 800 C
...........................................................................

"D. Discussion

Using a strong, reliable overhead stirrer is critical for this procedure"

"Using" -> "The use of" ?

[Edited on 24-2-2018 by Sulaiman]

aga - 24-2-2018 at 13:24

Is there some method to determine the Absolute absence of water that i've failed to encounter ?

Magpie - 24-2-2018 at 13:52

Quote: Originally posted by happyfooddance  

Why do you believe that a stirrer is critical? Dry boiling stones work for me.


The CaO/Ca(OH)2 slurry becomes very stiff. The CaO must come in intimate contact with the ethanol to take its water.

Magpie - 24-2-2018 at 14:04

Quote: Originally posted by Sulaiman  
In C,a,1 where did the aqueous ethanol come from and what is its concentration ?



This is a good question. There is a recipe. I need to look it up, again.
I keep this solution in a used alcohol can for reuse. It works really well when heated. It eats a very thin glass layer off the glassware.
Quote: Originally posted by Sulaiman  

In the writeup you set muffle furnace to 300 C but in a reply you mention 800 C


The 300C ifs for the mole sieves activation, the 800C is for converting Ca(OH)2 to CaO.
Quote: Originally posted by Sulaiman  

"D. Discussion

Using a strong, reliable overhead stirrer is critical for this procedure"

"Using" -> "The use of" ?


I don't understand this comment. Please explain.

[Edited on 24-2-2018 by Magpie]

Edit: Fixed broken quote

[Edited on 2-26-2018 by Texium (zts16)]

Magpie - 24-2-2018 at 14:06

Quote: Originally posted by aga  
Is there some method to determine the Absolute absence of water that i've failed to encounter ?


Yes, use of the Karl Fischer apparatus. http://www.cscscientific.com/moisture/karl-fischer

[Edited on 24-2-2018 by Magpie]

aga - 24-2-2018 at 14:22

I hate to be the Devil's advocate here, but did you perform a Karl Fischer titration on the resulting ethanol ?

Following procedures is great, but the older they are, the less likely that the result is actually 100% what you want.

Not being able to test the resulting material is something that seriously holds me back from doing many things.

I'm still trying to extract vicine from fava beans after a year or more.

Even when i get pure white crystals using a published paper, i have no way to know for certain what they are - might be just a sugar.

Magpie - 24-2-2018 at 14:46

No, I do not and never will, have Karl Fisher capability. I say my ethanol has <100ppm water because I followed the Vogel and Journal of Organic Chemistry procedures. Then I use my alcohol and get an excellent yield. This is good enough for me.

aga - 24-2-2018 at 14:54

Well, that makes it all OK, obviously.

A Chemist with your experience knows best, clearly.

Testing things is a waste of time - just follow something written years ago and it'll all be fine.

Magpie - 24-2-2018 at 15:00

Quote: Originally posted by aga  

Testing things is a waste of time - just follow something written years ago and it'll all be fine.


No, that's not the point. It worked as advertised. For every mole of water in your product you lose a mole of product. If your product has a MW of 180 you will lose 180g of product for every mole (18g) of water in your reaction vessel. If your ethanol is not dry you will get a lousy yield.

KF apparatus were $16,743 in 2000. If you would buy me one or rent time on one I will use it. It won't get me a better yield.

[Edited on 24-2-2018 by Magpie]

[Edited on 26-2-2018 by Magpie]

happyfooddance - 24-2-2018 at 15:08

If you make a concentrated copper sulfate solution and soak some strips of paper in them, and then dry them (until white obviously), they make a pretty good test strip for water, and will at least guarantee 99%, I am fairly sure of that.

Magpie - 24-2-2018 at 15:14

Here's the recipe for making the ethanol-water-NaOH glassware cleaner:

https://www.inrf.uci.edu/wordpress/wp-content/uploads/sop-we...

Sulaiman - 25-2-2018 at 04:29

Quote: Originally posted by happyfooddance  
If you make a concentrated copper sulfate solution and soak some strips of paper in them, and then dry them (until white obviously), they make a pretty good test strip for water, and will at least guarantee 99%, I am fairly sure of that.


Would Cobalt Chloride be a more sensitive water indicator than Copper Sulphate ?

S.C. Wack - 25-2-2018 at 11:34

Instead of CoCl2 one might prefer something insoluble. Like KMnO4.

Sulaiman - 25-2-2018 at 12:57

Quote: Originally posted by S.C. Wack  
Instead of CoCl2 one might prefer something insoluble. Like KMnO4.


Would KMnO4 not be immediately reduced by the ethanol, thus be no use as a water indicator ?

aga - 25-2-2018 at 13:13

Doesn't seem to happen at any appreciable rate even if it does.

In 100% ethanol it kinda just sits there as a crystal. Zero colour change.

With even a tiny amount of water it starts to dissolve, turning the liquid purple around the crystal.

AvBaeyer - 25-2-2018 at 16:02

Magpie,

Your procedure is interesting but seems like a lot of work to get a relatively small amount of anhydrous ethanol per unit of effort. Moreover, I doubt many here have access to a muffle furnace. Years ago when I made an honest living, we had a project that required the formation of a water sensitive imine. The reaction was run in commercial absolute ethanol using molecular sieves and worked fine except that it was a bit slow. When the sieves were activated in a muffle furnace like you did, the reaction was faster but there was no effect on the yield and thus not worth the extra effort required. Arguably, the situation is different if you are going to make, say, sodium ethoxide as a bit of water just plays havoc with yield. Again, we started with commercial absolute ethanol for this type of reaction but distilled it from diethyl succinate and a bit of sodium. Any sodium hydroxide formed from any water present caused the hydrolysis of the diethyl succinate to form a non-distillable sodium succinate salt. I think that this procedure is also in Vogel. The ethanol distilled in this manner gave reproducible high yields from sodium ethoxide based chemistry.

Nonetheless, I admire your work here as it solved a problem you needed to have solved. Good job.

AvB

happyfooddance - 25-2-2018 at 16:41

In lieu of a muffle furnace, a steel can will work, with a charcoal fire (a little air to the fire helps)

Magpie - 25-2-2018 at 16:48

Thank you, AvB.

I first learned about how water can kill yield when alkylating ethyl malonate per Brewster (forum library.) Brewster tells how to prepare dry ethanol. He starts with commercial absolute alcohol, which I've never had. (I now make an equivalent from Everclear [95% ethanol] and CaO.) Then he adds sodium, which I didn't have at the time. Then he adds diethyl phthalate and heats for 30 minutes. Again, I did not have this reagent then. Then he distills the ethanol into a bottle protected with a CaCl2 guard tube. The result is anhydrous alcohol. I don't remember how I dried my alcohol at that time but clearly it was inadequate as my yield was lousy.

Magpie - 25-2-2018 at 16:51

Quote: Originally posted by happyfooddance  
In lieu of a muffle furnace, a steel can will work, with a charcoal fire (a little air to the fire helps)


Actually 650°C will likely be adequate to to dehydrate Ca(OH)2, IIRC. Look it up on Wiki.

Edit: Wiki says Ca(OH)2 dehydrates at 515°C.

[Edited on 26-2-2018 by Magpie]

PirateDocBrown - 25-2-2018 at 18:24

Quote: Originally posted by Magpie  

silicone oil for oil bath (DOT 5 brake fluid works well)


That's a really good idea!

PirateDocBrown - 25-2-2018 at 18:31

Also, back in grad school, I used NaOH-saturated isopropanol (probably 99%, but I'm sure 91% would do) as a glassware soak.

You can't leave it in for more than a day or two as the NaOH would start to etch the glass. But it got almost everything clean, and far less dangerously than piranha sauce.

Magpie - 25-2-2018 at 18:35

I use hardware store denatured alcohol for the cleaning soln.

Edit: This may or may not be cheaper than drugstore 99% isopropanol. I will check.

[Edited on 26-2-2018 by Magpie]

[Edited on 27-2-2018 by Magpie]

Rosco Bodine - 4-4-2018 at 21:17

Using aluminum amalgam as a desiccant should be a lot easier. Cut some aluminum electrical wire into small rods, sufficient weight calculated to react with all the water present in the everclear plus maybe 2 or 3 percent extra. Degrease and etch the aluminum pellets with a 10% NaOH solution then rinse and blot dry. Drop the aluminum pellets into the everclear and add a couple of grams of mercuric chloride. The aluminum pellets will become covered with an adherent coating of metallic mercury which attacks and dissolves the surface of the aluminum and the resulting surface amalgam is highly reactive to oxygen, moisture, and alcohols, all being insoluble solid byproducts in this case with the alcohol being ethanol. The everclear should be completely dehydrated simply on standing for awhile, but stirring should speed the dehydration. Pressure relief should be provided for any free hydrogen byproduct. The weight of a solid glass stopper is sufficient as a check valve.

The dried alcohol could be distilled or simply filtered if a bit of AlCl3 is a tolerable impurity. If the amalgamation is done as advance preparation in a separate small amount of everclear, after the aluminum pellets are well amalgamated and shiny, the everclear having slight ALCl3 byproduct and any unreacted HgCl2 can be decanted and the "clean" amalgamated aluminum pellets used for the dehydration of the everclear.

The elemental mercury can be recovered by dissolving the residue in HCl and decanting the solution from the globules of mercury.

Absolute Ethanol should be the result of dehydration by aluminum amalgam .....meaning less than 1 ppm H2O. It might even serve well to store the absolute ethanol over aluminum amalgam to absorb any atmospheric moisture intrusion in storage. This would be a short term reaction for the aluminum amalgam because the aluminum would gradually react with the ethanol to form the insoluble aluminum alcoholate, Aluminum Triethanolate or Aluminum Triethoxide, but that would react with any incoming moisture to reform ethanol and insoluble aluminum hydroxide byproduct.

The 2010 JOC article on efficiency of various desiccants I did not read so I don't know if aluminum amalgam was tested. The idea I described is purely off the top of my head. I know aluminum amalgam is highly reactive and will even deoxygenate air as well as dehydrate the air.

[Edited on 4/5/2018 by Rosco Bodine]

Boffis - 5-4-2018 at 02:41

Can blue self-indicating silica gel be used to at least absorb some of the water from ethanol. Say to drop the water from 5% to say 1% first?

LearnedAmateur - 5-4-2018 at 03:26

Should be able to, I’m skeptical about 99% purity from a single drying run but you should be able to knock it down to a couple of percent. Just watch for that indicator change and if you use enough ethanol, you could probably dry the silica gel and distill off the water to figure out how much is removed.

Boffis - 5-4-2018 at 06:39

@LearnedAmateur, actually I was thinking that if I could drop the water content below even 2% water then I could azeotropically dry it by simply distilling of the more volatile azeotrope and leave essential dry alcohol in the still. The distillate should approximate to the azeotrope (95%) leaving a residue in the flask of purer alcohol which can then be dried with say 3A zeolite and finally distilled again.

RogueRose - 5-4-2018 at 12:00

Quote: Originally posted by Boffis  
Can blue self-indicating silica gel be used to at least absorb some of the water from ethanol. Say to drop the water from 5% to say 1% first?


I found it didn't even work with 90% or 180proof. Allowed to sit (while heated to 130) for 24 hours. Same % but lost about 15% due to adhesion.

aga - 5-4-2018 at 14:24

Why is there resistance, yet no follow-up experimentation, to prove or disprove deltaH's tripotassium phosphate hypothesis ?

All i did was throw KOH into H3PO4 to make some, and then added it to dilute ethanol and measure the result.

The result (in my experiments) was 100% ethanol, no water.

That could be +/- a few percent, yet to date, nobody has Bothered trying it, preferring to multum loqui without doing a damned thing.

What kind of Scientists are here these days ?

[Edited on 5-4-2018 by aga]

RogueRose - 5-4-2018 at 14:38

Quote: Originally posted by aga  
Why is there resistance, yet no follow-up experimentation, to prove or disprove deltaH's tripotassium phosphate hypothesis ?

All i did was throw NaOH into H3PO4 to make some, and then added it to dilute ethanol and measure the result.

The result (in my experiments) was 100% ethanol, no water.

That could be +/- a few percent, yet to date, nobody has Bothered trying it, preferring to multum loqui without doing a damned thing.

What kind of Scientists are here these days ?


Are you saying that you did this with Tri Sodium Phosphate, and DeltaH was suggesting TPP? Could you point to the DeltaH post or reference? I couldn't find a thing about anything you are talking about.

aga - 5-4-2018 at 14:45

use Google. Search like this :
site: sciencemadness.org ethanol tripotassium phosphate delta H

First result:
http://www.sciencemadness.org/talk/viewthread.php?tid=61746&...

Edit:

Trisodium phosphate remains untested. Nobody has reported on that.

I edited the post to put KOH where i said NaOH.

If You were to try the Na3PO4 experiment, then we'd all know something new.

[Edited on 5-4-2018 by aga]

RogueRose - 6-4-2018 at 04:43

Quote: Originally posted by aga  
use Google. Search like this :
site: sciencemadness.org ethanol tripotassium phosphate delta H

First result:
http://www.sciencemadness.org/talk/viewthread.php?tid=61746&...

Edit:

Trisodium phosphate remains untested. Nobody has reported on that.

I edited the post to put KOH where i said NaOH.

If You were to try the Na3PO4 experiment, then we'd all know something new.

[Edited on 5-4-2018 by aga]


I see. I just figured that you were talking in reference to this thread since you posted in this thread, kind of makes sense, otherwise people usually reference the thread they are talking about or at least mention it is a different thread.

RogueRose - 6-4-2018 at 14:31

Quote: Originally posted by aga  
use Google. Search like this :
site: sciencemadness.org ethanol tripotassium phosphate delta H

First result:
http://www.sciencemadness.org/talk/viewthread.php?tid=61746&...

Edit:

Trisodium phosphate remains untested. Nobody has reported on that.

I edited the post to put KOH where i said NaOH.

If You were to try the Na3PO4 experiment, then we'd all know something new.

[Edited on 5-4-2018 by aga]


Well I think TSP is an excellent suggestion to try and dry ethanol as if the Wiki is correct, it is insoluble in ethanol and the hydrate is a dodecahydrate, which I believe is 12 water molecules, so that suggests a strong drying capacity per mole.

There is no weight per mole listed in wiki but I calculated it as follows:
Na = 23 * 3
P = 31
O = 16 *4
Anhydrous = 164g/mole
12 molecules of water = 216g so dodecahydrate = 380g/mole

So theoretically this should dry 132% of weight in water that is added to the ethanol, so every 100g added to ethanol, 132g of water should be able to be absorbed.

I can try this but my TSP has some contaminates of Na2CO3 which shouldn't be much of an issue, it is only slightly soluble and is about 5% of the TSP. I'm curious if mine is the hydrate or anhydrous, I'm guessing hydrate as the anhydrous would be about 2.3x the price for same amount.

RogueRose - 6-4-2018 at 19:16

I have 200g TSP drying now and it's being a bugger. I have it in a glass dish and it has formed a hard dry layer on top and there looks to be a liquid/gel layer on the bottom that will just not go away even with the oven on "Toast" (both bake/broil at same time - on high). It looses water at 212 so the oven at 525 should be more than adequate at 3 hours+. That anhydrous MP is crazy high near 3000F so that layer isn't melted anhyd TSP. There is some Na2CO3 in it but the MP is near 1500F, so it isn't that either. IDK how long this should take or a better way to do this w/o removing and chopping it up to re-heat.

aga - 7-4-2018 at 00:10

Yes, it does like it's water, which is what makes it so effective i guess.

A 1931 industrial process for Na3PO4 ends with .12H2O hydrate crystals, and starts with neutralising the phosphoric acid to disodium phosphate first, mainly to remove impurities from the acid it seems, then to trisodium phosphate at the end.

Sounds like a good way to go if your acid is low grade and green like mine.

This sounds do-able, despite mentioning a 4,000 gallon tank and a 70 foot tower :o

Attachment: snell1931.pdf (755kB)
This file has been downloaded 859 times

Boffis - 3-6-2018 at 03:27

@ Magpie; Is it necessary to heat the suspension of calcium oxide in the alcohol? Could I simply leave the suspension, perhaps with occasional shaking, for a week or so and then let it settle, decant or filter of the alcohol and distill? This way I don't need the "powerful overhead stirrer" which I don't have. Does the ethanol-calcium oxide mixture get hot when you first mix them? Isopropanol azeotrope and calcium oxide certainly do.

I tried this method with recovered isopropanol and the calcium oxide set into a concrete-like block and then I just pour off the isopropanol and redistill it with a high reflux ratio and low distillation rate. I do this in old plastic jerry cans and just throw them away when I am done. I actual fact I will probably not do this again for isopropanol because it is now so easy to get at 99%+ and so cheap it isn't worth it, particularly as the azeotrope is about 30% water.

Magpie - 3-6-2018 at 06:09

Quote: Originally posted by Boffis  
@ Magpie; Is it necessary to heat the suspension of calcium oxide in the alcohol? Could I simply leave the suspension, perhaps with occasional shaking, for a week or so and then let it settle, decant or filter of the alcohol and distill? This way I don't need the "powerful overhead stirrer" which I don't have. Does the ethanol-calcium oxide mixture get hot when you first mix them? Isopropanol azeotrope and calcium oxide certainly do.


I don't think this would work.

Boffis - 3-6-2018 at 12:28

@Magpie; Why not? Any actual technical reason or just a gut hunch? As I say the reaction between azeotropic isopropanol and calcim oxide is exothermic, not as exothermic as with water but still the mixture gets hot.

What difference does the heating do apart from speed up the reaction? If its an equilibrium reaction surely low temperatures favour calcium hydroxide though I can't see there being much difference between room temperature and 78 C making much difference.

Anyway I have set up an experiment with 500ml of rectified spirit and 81g of calcium oxide and I see how it goes.

Magpie - 3-6-2018 at 15:01

The reaction between CaO and water is 2-phase. Therefore, strong mixing is important. Even with strong mixing I call for 6 hrs of mixing. Check Vogel.


Edit: On checking Vogel I see that he does not prescribe the use of a stirrer, but refluxing only. I'm concerned that one might have severe bumping without a mixer, however.

[Edited on 3-6-2018 by Magpie]

walruslover69 - 4-6-2018 at 06:05

Over the summer I have access to a piece of spectroscopy equipment that analyzes the density and specific gravity of solutions to the 0.000001 g/cm and is used to determine compositions of binary solutions. If you guys commented or messaged me with a list of discussants you would like to test for their drying abilities, I would love to be able to make a table with the drying properties of desiccants that everyone on the forum could reference.

should I make this project a separate thread or leave it in here?