Sciencemadness Discussion Board - Hydroxyacetone via dehydration of glycerol

Hydroxyacetone via dehydration of glycerol

Assured Fish - 26-1-2017 at 18:41

HI I have recently managed to obtain about 100g of some copper chromite from an ebay supplier. From what the product number says it was purchased from Sigma Aldrich originally.

http://www.sigmaaldrich.com/catalog/search?term=209317&i...

I want to try out a dehydration reaction using this to produce hydroxyacetone from glycerol as described in the following patent.

https://www.google.com/patents/EP2024321B1?cl=en

In the examples down near the bottom the patent describes placing glycerol of varying purity in a RB flask with a gas inlet and a dean stark trap followed by heating to around 230*C with stirring and gas sparging with N2 gas.
I want to try doing this with a standard distillation set up without the sparging and using heating tape and a variac to maintain the temperature at 230*C and of coarse accompanied with strong magnetic stirring.
I think i will be starting with 100g of USP grade vegetable glycerin and 2.8g of copper chromite, then further purifying via fractional distillation and perhaps drying over activated calcium chloride or another drying agent; I will be using a 200ml column for this but im a little afraid that the crude product might get to hot and perhaps decompose a little wile doing this after all it has a boiling point of 145*C and the pot will be considerably hotter than that.
My question is how necessary would the N2 sparging be for this reaction beyond simply increasing the yeild, I do not see why an inert atmosphere would be necessary but im rather inexperienced with making such a prediction.

Cheers Fish

Praxichys - 27-1-2017 at 06:40

The chromite catalyst and high heat will probably cause the hydroxyacetone to undergo an aldol condensations, condensing it into goo. This goo will probably trap further hydroxyacetone due to polarity and viscosity.

The hydroxyacetone may not be formed quickly enough to escape the reaction by its own vapor before it refluxes back into the pot and is destroyed, hence the nitrogen sparging to carry it. Without the sparging, there could be little to no yield. However, the authors may have simply been trying to test the mechanistic efficiency and therefore designed the experiment so that the products were removed as quickly as possible for quantitative analysis. It could still have a decent yield without the sparging - it's hard to say without trying it.

As for the nitrogen, I can't see why ordinary dry air could not be used, other than that it might induce a few competing oxidation reactions. It could also just be that the authors didn't want to introduce further variables. Glycerol is cheap; try it and let us know!

Metacelsus - 27-1-2017 at 12:48

Watch out for acrolein. That stuff is quite nasty.

tsathoggua1 - 27-1-2017 at 12:49

Is it not more likely for acrolein to be the product? Thats what tends to happen when dehydrating glycerine. Or does the copper and/or Cr participate in a different manner to the usual methods of dehydrating glycerine, such as with Na/K bisulfate or thermolytic dehydration?

High temperature and glycerine seem like the result, if not polymeric tarry crap in the first place, would form at least some acrolein, which would then itself probably undergo aldol condensation. Hopefully before any gets out of the reaction flask, since it is a rather unpleasant chemical to be around. And not terribly stable itself.

Edit-metacelsus has the same idea and posted at the same time, I see.

Nasty doesn't quite cover it. Its a powerful lachrymator, it smells foul and its pretty toxic stuff.

[Edited on 27-1-2017 by tsathoggua1]

Assured Fish - 27-1-2017 at 18:04

Quote:

The chromite catalyst and high heat will probably cause the hydroxyacetone to undergo an aldol condensations, condensing it into goo. This goo will probably trap further hydroxyacetone due to polarity and viscosity. The hydroxyacetone may not be formed quickly enough to escape the reaction by its own vapor before it refluxes back into the pot and is destroyed, hence the nitrogen sparging to carry it. Without the sparging, there could be little to no yield. However, the authors may have simply been trying to test the mechanistic efficiency and therefore designed the experiment so that the products were removed as quickly as possible for quantitative analysis. It could still have a decent yield without the sparging - it's hard to say without trying it. As for the nitrogen, I can't see why ordinary dry air could not be used, other than that it might induce a few competing oxidation reactions. It could also just be that the authors didn't want to introduce further variables. Glycerol is cheap; try it and let us know!

Yea the acetol refluxing back into the flask may be problematic especially due to me not having a 3 or 2 neck flask on hand and instead i will be using a claisen adapter, hopefully this won't be too problematic if I use some insulation. Like you said however i guess i will just have to try it and find out.
Your idea about using straight air might be a good option if things with the first experiment don't go so well, i need to get myself a thermocouple and then i could make the claisen adapter system work with a gas inlet tube.

Quote:

Is it not more likely for acrolein to be the product? Thats what tends to happen when dehydrating glycerine. Or does the copper and/or Cr participate in a different manner to the usual methods of dehydrating glycerine, such as with Na/K bisulfate or thermolytic dehydration? High temperature and glycerine seem like the result, if not polymeric tarry crap in the first place, would form at least some acrolein, which would then itself probably undergo aldol condensation. Hopefully before any gets out of the reaction flask, since it is a rather unpleasant chemical to be around. And not terribly stable itself.

According to this diagram the glycerol must first be dehydrated to 3-hydroxyproponaldehyde before being further dehydrated to acrolein which could undergo decarboxylation before a further dehydration to produce other products thus making acrolein if formed at all a rather minor product but i will be doing this reaction in a well ventilated place with a dust mask which would atleast prolong inhalation hopefully giving me a warning with a lacromating effect insetad.
In the case where acrolein does form as a precaution i could set up a hot water bath on the receiving flask and an inverted funnel off the vacuum takeoff adapter however im not sure what would be the best option for neutralizing the acrolein, what would you suggest?

Praxichys - 27-1-2017 at 18:38

You can scrub acrolein easily with a cold, strong NaOH, KOH, or sodium bisulfite solution combined with an efficient gas/liquid contact scheme (like a fritted sparging wand or a capillary tube/rapid stirring). In the case of the strong hydroxides, the solution will cause the acrolein to self-condense which will be obvious by yellowing of the solution. In the case of the bisulfite, a soluble aldehyde bisulfite adduct will form.

tsathoggua1 - 28-1-2017 at 04:25

If not scrubbed, even as a minor product, without a mask and eye protection (one could just use a blast shield to protect the face, rather than pairing it with goggles to keep the fumes from the eyes, but still protect from any splashes, and do use a respirator mask)

If its there at all, you'll know about it, at least unless in the tiniest of traces) It is not pleasant, and the fumes, as stated, are vicious. If I had to make the comparison, I'd say its an awful lot worse than chloroacetone.

Success

Assured Fish - 7-2-2017 at 21:51

To a single neck 500ml RB flask was added 2.825g of copper chromite followed by 100g of glycerol, to the round bottom flask was attached a still head and a thermometer was put into the thermometer adapter so as the bulb of the thermometer was immersed in the reaction mixture. A 200mm liebig condenser was then attached to the still head and a bent vacuum takeoff adapter and a 250ml RB flask was attacked to collect the distillate.
As a cautionary measure a tube was attached to the vacuum connection on the receiver and this tube was lead outside i case of the formation of acrolein as a byproduct.

The 500ml RB flask was then heated with stirring so that a small vortex was created but the bulb of the thermometer still remained submerged.
At the beginning the stir bar was unable to move however when the solution temperature reached above 50*C this was no longer an issue.

The reaction mixture was slowly heated to 230*C during which time the reaction mixture turned black and murky and a small amount of what looked like water vapor mystified the RB flask (this was likely just residual water from the glycerin).
At around 230*C a vapor front was seen climbing up the still heat however the RB flask and still head were insulated using aluminium foil in order to help the front move into the condenser.
At first distillation proceeded extremely slowly with only about a drop every minute or so however i ramped the temperature up to around 240*C and started collecting at around 1 drop every 20 seconds or so. I found that keeping the temperature at around 245*C yielded a drop every 10 seconds and i kept it as this however i did get the heat to just over 250*C with a noticeable increase in the collection rate, I never let the reaction temp get above 255*C.

There was seen a yellow translucent liquid smelling strongly of caramel in the collection flask.
The distillation was stopped at around 6 hour however by this time the collection rate was still going and the distillation could have been continued for at least another 3 hours by my estimate. the reaction mixture by this time looked like a hot chocolate

.
Foolishly i forgot to weigh the collection flask so i had to transfer to a beaker but the yield was 39.2g however it should be noted that this product is contaminated with large amounts of water and i will need to further distill it to obtain pure hydroxyacetone, i will be doing this in the next few days and will post the final yield when i can.

Notes:
*Throughout the distillation no lacramatory effects were detected at all however given that glycerol decomposes to acrolein above 280*C one should be careful monitoring the reaction temp but i suspect the temp would be fine to keep at 260*C.
*When i added the glycerol to the copper chromite in the flask the mixture took on a very viscous characteristic and i suspect this is due to polymerization as Praxichys had stated and this would have slowed down the reaction and reduced yields and this im guessing is why the original patenter used gas sparging which would significantly speed up the reaction.
But despite this it is still very possible to produce hydroxyacetone via this method without gas sparging but the distillation will take a wile, i speculate 8 hours for a 100g run of glycerin however i stand by the fact that this reaction may be able to produce up to 70% yield and given how cheap the reagents are this is in my opinion an excellent means of producing hydroxyacetone but i will recomend using gas sparging to anyone wishing to recreate this.
*The supposed hydroxyacetone itself had a very unexpected odor smelling like sweet but burnt caramel and was sickly when smelt closely. I then found some information regarding its physical characteristics hense my confidence in this being successful however i suppose i cannot truly know for sure without doing a distillation to deturmine the boiling point. http://www.thegoodscentscompany.com/data/rw1132561.html
*It would be fucken epic if someone were to carry out an esterfication between hydroxyacetone and phenylacetic acid as im sure the resulting compound would having a very interesting aromatic profile :cool: