[CaptainPanic]

This is a (conceptual) engineering problem (not homework, I'm not in school anymore)... And I thought I'd entertain you all with it. Perhaps it should be moved to engineering, perhaps it should be here in applied chemistry.

The problem: there is a stream of 5% glycerol () and 94% water, 1% salts. The flow is 5 kg/s (large). Temperature is 99 deg Celcius. The whole stream is in the liquid phase. Pressure is atmospheric.

How to obtain a clean glycerol stream from this at low cost (cost effective, glycerol is the only product of any value)?

I have considered:

distillation (not very elegant in this case, salt remains in the glycerol)

extraction (viable option, i.e. with hexane) [edit] no, hexane is too non-polar. Glycerol and biodiesel (RME) form two phases, so I don't believe that hexane is the right solvent. Something more polar needed :D

membrane separations (glycerol selective membrane)

and crystallization (no clue if that works)

Has anyone heard of a glycerol-water separation? I'll be looking into it, so if I find something, I'll post it. This is no "I'm completely lost, help me please - thread"... but just for your and my entertainment on Friday afternoon. Brainstorm away, silly ideas are also welcome :D

p.s. a solid-liquid phase diagram for the glycerol+water system would be welcome. :D

[YT2095]

what do you want to isolate glycerine for? making some GTN from OTC medicines are we???

[John Cuthber]

I don't know where you buy stuff YT2095, but I don't think of "The flow is 5 kg/s (large). " as being OTC.
My guess is a byproduct from biodiesel making.
I'd think about vacuum stripping a lot of the water out then desalting it by ion exchange then vacuum distillation to clean it up.
Freezing the stuff might strip out some of the water, but if its already near boiling that would seem clumsy.

Another aproach might be to find a use for the dilute, impure stuff. That's the holy grail for biodiesel makers at the moment.
If, for example, you could find some bacterium/ yeast that could ferment the glycerol to ethanol, or methanol you could use that for the esterification which would be very neat.

[CaptainPanic]

YT2095, thanks for your concerns. (I really appreciate that you try to keep this forum decent). Glycerol is a by-product of a micro-organism also makes ethanol (a GMO yeast).
There is no actual factory, and also no clue what to do with loads of glycerol (sell it, to a fictive buyer in my model)... :D

We try to make a few extra euro's per tonne of ethanol by finding a way to separate the glycerol too. But the glycerol market isn't very good at the moment (because of all the biodiesel factories that also have it as a by product) so it must be cheap. It's a theoretical new yeast that actually shits the glycerol out, so I doubt it's interested to eat it again. :-p

I've given up on all 4 options for the following reasons:

Distillation leaves all solids and salts in the glycerol phase... Ion exchanging and filtering is an option, but it all becomes a bit too expensive.

I couldn't find any decent solvent for extraction. It either forms a single phase (e.g. ethanol, acetone), it dissolves more water than glycerol (e.g. MIBK), or it doesn't do anything at all (e.g. hexane, chloroform). I've modeled it in a flowsheet program using the UNIQUAC method, so only approaching reality, not actually testing it.

I have no info on glycerol selective membranes. Pushing water through the membrane is pointless, because like with distillation all the other stuff (salts, solids) stay with the glycerol.

Glycerol and water seem to form a single phase when crystallized (http://www.dow.com/g...rces/table8.htm).

We're investigating now if a digester would work on glycerol. Google gives a few results, so we might convert it to swamp gas (methane, CH4).

[hermanntrude]

perhaps you could use a MORE polar solvent to isolate it? an ionic liquid or something?

or perhaps you could use the opposite of a chelating agent? add a transition metal ion to which only glycerol can bind... since glycerol is a tridentate ligand and water is only monodentate... that shouldn't be too troublesome

OR, perhaps you could react it so it precipitates out?

just random ideas

[YT2095]

I wonder if Hydrogenation would work to convert it to Propane?
maybe Pyrolysis of the soln to generate the hydrogen?

[hermanntrude]

YT2095 said:

I wonder if Hydrogenation would work to convert it to Propane?
maybe Pyrolysis of the soln to generate the hydrogen?

sounds a bit heavy-handed to me. I think the desired result is to actually get glycerol back again. I was thinking of a small, simple reaction which can be easily reversed. Just a method for separating the glycerol

[YT2095]

how about Esterification and Hydrolysis then?

[hermanntrude]

perhaps.

My first try would be to form an alkoxide, since that'd be ionic and so easier to precipitate out

[John Cuthber]

Bit tricky in the presense of a lot of water.

[hermanntrude]

ah. yes. good point. There's always something isn't there?

I say chelating is best then. Add a metal which will accept glycerol as a ligand in preference to water then isolate the coordination compound thus formed.

[CaptainPanic]

Thanks all for replying. Random thoughts are very nice, and sometimes they're brilliant :D

For chelating, shouldn't the glycerol form an ion as well? The ligand in EDTA chelation for example is the 4- ion. I think that the equilibrium in the case of glycerol will not be favorable to form any negative ions (an group doesn't like to form an ). Water is a powerful ligand, especially since it stabilizes itself with neighbouring water molecules.

I like the idea of ionic liquids. There are ionic liquids that are insoluble in water. My only fear is the costs of such a operation. There will always be a small loss of ionic liquid into the water stream when you use it in an extraction... so there is always a need to have an ionic liquid make-up stream, which could be immensely costly.

Perhaps esterification isn't such a bad idea. I cannot think of any valuable esters that include glycerol (other than vegetable oils, which would be kinda pointless), but perhaps I'll think of some in the future.

I'm still looking at digestion. I found an article (click, then click on "full text") about a microorganism that converts glycerol to ethanol. It is from 1999 and no follow ups were published (none that I could find), so I am not sure how good this works. No yield was published in the article.

Glycerol can be used in a normal digester too (forming methane as product). Not exactly rocket science, but it is perhaps the most feasible option.

[hermanntrude]

CP i was thinking of glycerol acting as a ligand by hydrogen bonds, not by ionic bonds. I figured since it'd be tridentate rather than monodentate like water, it'd bind more strongly

[CaptainPanic]

I don't think that they are close enough. The -OH groups don't seem to point at one point. If they'd all three point towards a central position I think it could work.

[hermanntrude]

glycerol is definitely a ligand, and from the model I just made i'd say it's perfectly capable of being tridentate

[John Cuthber]

IIRC there are speciality HPLC columns that are used for separating sugars which rely on boronic acid groups bonded to the mobile phase.
Glycerol definitely forms some sort of complex with boric acid and this might be employed in some way, like the HPLC columns, as a way to extract the glycerol.

[hermanntrude]

yes, boronic acids form complexes with diols

[John Cuthber]

Thanks for pointing out that I wasn't lying and that the people who make those HPLC columns are not idiots.

[hermanntrude]

that's OK. No problem. Any time. :)

[a touch of electricity]

here's a light bulb moment.
if you do a 2 stage process, electrolysis can separate the water into hydrogen and oxygen without effecting the glycerol.
here's why:
for electrolysis separation to occur, the substance you wish to separate must be an electrolyte.
water is an electrolyte, but glycerol (aka glycerin) is not.
so the water will turn to gas while the glycerol stays in it's molecular form.

how to separate the salts is, to me, undiscovered.

however, if you need an extremely dry, air tight chamber for testing electrostatic experiments, this is simple and cheep.
it only takes a few but no more than 12 volts DC to do so.

car battery anyone?

if you'd like, the gases released from this process are hydrogen and oxygen, so if you filter out the hydrogen and fuel your hydrogen fuel cell car then you're really got it good!