Sciencemadness Discussion Board

Separation of Xylenes using Formic Acid Azeotrope

Meltonium - 28-9-2016 at 06:43

Xylenes commonly are found in the hardware store sold as Xylol. It comes as a mixture of the o, m, and p isomers. Sometimes it is necessary to separate these isomers which can be hard.
By adding formic acid to the mixture an azeotrope forms between it and the m-xylene that boils at ~97C. M-xylene is not used for much so it can be kept in with the formic acid. The rest of the mixture then boils off at ~140C as a combination of o-xylene and p-xylene. The boiling points of these are similar but their freezing points aren't. P-xylene freezes at 13C and can be extracted. O-xylene freezes at much lower temperatures and will remain a liquid. From there you have separated isomers.

I have not actually validated this procedure. Please tell me if there is something wrong with this. If nothing is wrong then I recommend going through the procedure a few times especially the crystallization of p-xylene to obtain a purer product.

[Edit] Clarified title

[Edited on 9-28-2016 by zts16]

Sulaiman - 28-9-2016 at 08:31

just a noob thinking ....
could the m-xylene be separated from o-xylene and/or p=xylene by freezing ?
if so the overall setup would be simpler.
Doing the formic acid separation after separation by freezing would give at least one isomer free of formic acid contamination.

Meltonium - 28-9-2016 at 08:39

Quote: Originally posted by Sulaiman  
just a noob thinking ....
could the m-xylene be separated from o-xylene and/or p=xylene by freezing ?
if so the overall setup would be simpler.
Doing the formic acid separation after separation by freezing would give at least one isomer free of formic acid contamination.


Technically it could as the freezing point of m-xylene is -25C while o-xylene is -45C but the mixture of all the Xylenes has a freezing point of ~-44C so I think by removing the m-xylene first the temperature difference between the FP of o and p xylene is enough to have the p-xylene freeze separately

Meltonium - 28-9-2016 at 08:40

Quote: Originally posted by Sulaiman  
just a noob thinking ....
could the m-xylene be separated from o-xylene and/or p=xylene by freezing ?
if so the overall setup would be simpler.
Doing the formic acid separation after separation by freezing would give at least one isomer free of formic acid contamination.


Technically it could as the freezing point of m-xylene is -25C while o-xylene is -45C but the mixture of all the Xylenes has a freezing point of ~-44C so I think by removing the m-xylene first the temperature difference between the FP of o and p xylene is enough to have the p-xylene freeze separately

Texium - 28-9-2016 at 08:42

To me the procedure seems plausible, and I would be interested in trying it sometime but I have a lot on my plate right now so it would have to go to the back burner along with many other procedures of interest. Why don't you try it out, Meltonium? Keep in mind too that commercial xylene may contain up to 20% ethylbenzene, which may complicate things. It's worth a try though.

Formic acid could easily be removed from any xylene isomer by extracting with a dilute sodium hydroxide solution and then drying/redistilling the xylene.

By the way, I changed the title of the thread because separation of xylenes has been discussed many times here before, but not by this specific method as far as I can tell.

[Edited on 9-28-2016 by zts16]

Maroboduus - 28-9-2016 at 09:25

Water forms a 94.5C boiling azeotrope with Meta Xylene as well, but that doesn't mean it is useful for separation. Still it would be cheaper to try than formic acid.
I suspect that there's just more azeotrope info for the meta than for the other isomers, and that it'll be a big mess when you actually try distilling the isomers with all sorts of azeotropes of the three isomers coming over.
I sure hope not, this would be really handy. I am skeptical however.

Same with the freezing thing. Freezing out to separate similar substances can be tricky because of eutectics and all that jazz.
And bear in mind that the freezing point is for the pure compound without solvents involved. Sucrose freezes at 186C, but you can't freeze the sucrose out of an aqueous solution by cooling it below 186C.

In my admittedly limited experience of purification by freezing, I have found that mixtures of similar components have a tendency to just sort of gel and slowly solidify. Sometimes you get better results if there's some solvent in there of a dis-similar character to the components you're trying to separate.



[Edited on 28-9-2016 by Maroboduus]

Meltonium - 28-9-2016 at 09:37

M-xylene also forms an azeotrope with acetic acid at around 117C and afterward it can be removed by water washings because Xylenes are practically insoluble in water. As for the azeotrope meta has with water, I didn't find that. Although my source is the azeotrope tables from Wikipedia. If there is ethylbenzene contamination, it could be removed through water washings because it is very slightly soluble.

Texium - 28-9-2016 at 09:48

Quote: Originally posted by Maroboduus  
In my admittedly limited experience of purification by freezing, I have found that mixtures of similar components have a tendency to just sort of gel and slowly solidify. Sometimes you get better results if there's some solvent in there of a dis-similar character to the components you're trying to separate.
While that is often true, sometimes it is not. For instance, p-nitrotoluene and o-nitrotoluene are separable by freezing out the p-nitrotoluene. I don't know if it would work in this situation, but removing the m-xylene, which is supposed to be about 40% of commercial xylene, should increase the likelihood that it will.

Boffis - 28-9-2016 at 10:37

Hi Meltonium, do you have a reference for this procedure? It sounds potentially very interesting but I wonder if the azeotrope with m-xylene only forms a simple system in the absence of the o and p-isomers so that with a mixture of the 3 isomers you get a "common" azoetrope. It would be very useful if it worked from the mixed isomer commercial xylol though :).

Also what strength is your formic acid; 99% or the standard 85% industrial grade?

Interestingly I have been experimenting recently with the azeotropic distillation of 85% formic in an attempt to upgrade it to >97% so I can use it directly to product formate esters. What is interesting is that instead of the water distilling over with the solvent the formic acid distils over in a slightly enriched form; 85 up to 93% in one distillation with petroleum ether (80-100 C boiling range). I found that the higher strength formic acid was not misible with the petroleum and so could be separated in a separating funnel. So if the m isomer distils off with the formic acid it should be easily separated.

Like you say the o and p isomers should be fairly readily separated by the freezing method. I have a reference to this somewhere amongst my papers. Another method of removing the m isomer is by treating with sulphuric acid. This works on the principle that only the m-isomer reacts with sulphuric acid of <85% and since water is generated by the sulphonation of xylene the liberated water soon limits sulphonation to solely the m-isomer, the o and p enriched residue can then be fractionated or frozen to separate thje remaining isomers. There are several refs to this procedure, I seem to remember that the original lead was from Ullmans encyclopedia.


Maroboduus - 28-9-2016 at 10:44

Quote: Originally posted by zts16  
Quote: Originally posted by Maroboduus  
In my admittedly limited experience of purification by freezing, I have found that mixtures of similar components have a tendency to just sort of gel and slowly solidify. Sometimes you get better results if there's some solvent in there of a dis-similar character to the components you're trying to separate.
While that is often true, sometimes it is not. For instance, p-nitrotoluene and o-nitrotoluene are separable by freezing out the p-nitrotoluene. I don't know if it would work in this situation, but removing the m-xylene, which is supposed to be about 40% of commercial xylene, should increase the likelihood that it will.


The polarity of a nitro group may make this more possible. My experiences are largely with essential oils which have less dramatic polarity, and more complex structures (lots of alkyl and phenol ether side chains).
I can see where, given the different resonance structures contributing to the character of the para isomer, it would be quite different from the other xylene isomers in charge distribution(not just it's symmetry, but more charge on the methyls), but I'd still bet against getting a good separation by cooling. But I do think it's worth trying.

EDIT: turns out there are plenty of patents that involve crystalizing out the Para isomer at -30C to -60C. These are often patents for detailed processes to separate the isomers which involve fractionating and freezing.

[Edited on 28-9-2016 by Maroboduus]

Meltonium - 28-9-2016 at 11:38

Quote: Originally posted by Boffis  
Hi Meltonium, do you have a reference for this procedure? It sounds potentially very interesting but I wonder if the azeotrope with m-xylene only forms a simple system in the absence of the o and p-isomers so that with a mixture of the 3 isomers you get a "common" azoetrope. It would be very useful if it worked from the mixed isomer commercial xylol though :).

Also what strength is your formic acid; 99% or the standard 85% industrial grade?



I have no reference other than myself and Wikipedia azeotrope tables. XD

As for the formic acid, I have none at the moment but I'd imagine that it would need to be near pure because the azeotrope it forms with water has a lower BP than the azeotrope formed with the m-xylene.

When I get around to doing this (hopefully this weekend if time allows) I will probably use glacial acetic acid instead of formic acid and let yall know how it goes.

Texium - 28-9-2016 at 11:46

I have 95% formic acid, so I can try it out some time and make a video of the procedure, but currently I have a few other projects that I'm working on that I'd like to complete first. If this thread ends up getting buried without anyone trying it, I will make sure to revisit it and post my results.

Meltonium - 1-10-2016 at 14:05

I've preformed the separation using acetic acid and below is my written work:

Separation of Xylenes


Objective: To separate xylene isomers using an acetic acid azeotrope

Materials:

- 200mL Xylenes
- 100mL Glacial Acetic Acid
- Distilled Water

Procedure:

1) Add the Acetic Acid and Xylenes to a 500mL RBF
2) Mix for ~10 minutes
3) Fractionally distill over at 117C
4) Wash distillate with water to obtain Acetic Acid and dry over a drying agent
~
5) Take remaining fractions that distill over 120C
6) Place fractions in a freezer for ~1-2hours
7) Decant solution and collect crystals in separate container
8) Repeat for purer product

Observations/Analysis:

- Upon addition of the Acetic Acid, the solution became cloudy. This is likely from a suspension of undissolved acetic acid or from water present in the acid.
- Upon heating, the solution underwent significant thermal expansion bringing it close to overfilling the RBF
- The first fraction started coming over at ~95C; the distillate separated into two layers, one cloudy and one clear. They both seemed to be growing at the same rate. As more distillate collects, the top layer seems to be clearing up and the bottom layer becoming more cloudy.
- A new fraction was collected when the temperature reached 111C. This fraction is clear and smells of xylene. The temperature stabilized at 115C for most of this fraction.
- A third fraction was collected when the temperature rose to 120C
- To prevent possible photo-degradation of the separated xylenes (as claimed in another post under phthalic anhydride from o-xylene) the lights were dimmed.
- A fourth fraction was collected when the temperature rose to 132C. This fraction stabilized at 140C into the fifth fraction.
- Distillation was stopped at the start of the fifth fraction; most of the liquid in the RBF had been distilled.
- Left in the RBF is a dirty, champagne colored liquid.
- Fractions 3-5 were transferred to a 250mL beaker, covered with Saran wrap, and placed in the freezer (-14C). Thin flaky crystals formed in the bottom of the beaker after ~1 hour.

Because the azeotrope of acetic acid and meta-xylene boils at 117C, it is reasonable to assume that fraction two is almost all acetic acid and m-xylene. Fractions 3-5 are most definitely a mixture of xylenes, ortho being most prominent after the p-xylene crystals are removed. Due to the lack of abundance of p-xylene crystals, it seems like the ortho and meta isomers are most prevalent in the unpurified xylene. There is likely ethyl benzene mixed with this o-xylene because the MSDS for the commercial Xylol used suggested 0-30% ethyl benzene. The ethyl benzene could be removed through many, many water washings as it is more soluble than o-xylene. As for the first fraction, I suspect that the cloudyness is from a suspension of water droplets that may have been present in the RBF, collection beaker, or the acetic acid.

About 5-10mL of p-xylene was collected after the crystals were remelted.

[Edited on 2-10-2016 by Meltonium]

Metacelsus - 1-10-2016 at 16:33

Those are definitely some interesting results. What kind of distillation column were you using?

Also, is there any good way to check the proportions of xylenes in your fractions? (GC would be obviously the best, but that's probably not available.) Could you do something with the density or refractive index?

Meltonium - 1-10-2016 at 16:40

Quote: Originally posted by Metacelsus  
Those are definitely some interesting results. What kind of distillation column were you using?

Also, is there any good way to check the proportions of xylenes in your fractions? (GC would be obviously the best, but that's probably not available.) Could you do something with the density or refractive index?


I used a 400mm Vigreux column connected to a 400mm Leibig condenser although a 300 leibig would've done just fine.

I was thinking about how to check the proportions, but I don't think density is the best way for me. I don't have a scale or a graduated cylinder with enough decimal places to differentiate between the isomers since they have pretty similar densities. I didn't think to look at the refractive index, but that's a good thought and I'm open to any more suggestions.

- -

Turns out the refractive indexes are pretty much the same too. :(

[Edited on 2-10-2016 by Meltonium]

Metacelsus - 2-10-2016 at 05:59

Maybe you could oxidize a sample of them to the dicarboxylic acids, and analyze those by TLC? They'll probably be visible as separate spots.

Meltonium - 2-10-2016 at 06:56

Quote: Originally posted by Metacelsus  
Maybe you could oxidize a sample of them to the dicarboxylic acids, and analyze those by TLC? They'll probably be visible as separate spots.


Good idea! I'll try that just as soon as I get this crud out of my 100mL RBF

Meltonium - 2-10-2016 at 17:02

I'm a little confused about the oxidation of these isomers and the contaminants. Organic chemistry is not my fortay, but I know the xylene isomers will be oxidized to their phthalic acids, but my fraction that I think contains mostly ortho-xylene is likely to have ethyl benzene as well. I do not know what ethyl benzene will oxidize to. Can yall help me?

P.S. I plan to use potassium permanganate to oxidize my samples if that helps any :)

Metacelsus - 2-10-2016 at 17:51

Ethylbenzene would be oxidized to benzoic acid by permanganate.

Maroboduus - 12-10-2016 at 13:56

So any idea what that first fraction was?

EDIT: And what were the volumes of the fractions?



[Edited on 12-10-2016 by Maroboduus]

byko3y - 12-10-2016 at 15:54

You need approx 70 theoretical plates to efficiently separate o-xylene and m-xylene (vs 100 for pure xylenes). I have no idea what heppens to other isomers. Formic acid is one of the best azeotropic entrainer for xylene isomer separation, yet the difference is not as large as you might wish.

Meltonium - 21-10-2016 at 15:25

Today I attempted to oxidize my ortho fraction and convert it to phthalic anhydride and use that to indicate the % of ortho really in the fraction. ~3 hours into the oxidation through KMnO4 I mistakenly grabbed a beaker labeled for HCl instead of distilled water and added that to the reaction. Needless to say, I scrapped what was left and won't be attempting it again any time soon. (It is very messy) So I'll probably send my fraction to a lab for analysis. When the results come, I'll scan the document and post it here.

On another note, I am very certain that the first fraction I collected was un-azeotroped acetic acid along with low-boiling organically with suspended water droplets.

Meltonium - 21-11-2016 at 09:38

Quote: Originally posted by Maroboduus  
So any idea what that first fraction was?
[Edited on 12-10-2016 by Maroboduus]


Previously, it was unknown to me that there was an azeotrope between ethyl benzene and water, as well as an azeotrope between m-xylene and water; both boil at 92C according to my source. The first fraction likely then was water azeotroped with the ethyl benzene or some m-xylene and when it condensed and cooled the liquids separated out.

Knowing this, I can now separate the ethyl benzene from my o-xylene fraction using the water azeotrope. Ethyl benzene is assumed to be the major contaminant of the o-xylene fraction.

On another note, both the m-xylene and the ethyl benzene can be separated from the original mixture using this water azeotrope. The BP is lower than that of the acetic acid azeotrope I used, therefore; the fractional distillation is likely to yield a purer product.

I will try the water separation method and post the results below.

Results

I used ~500mL of Xylol and from that ~350mL came over at 92C with the water added.
The still head then climbed to 100C briefly and progressed until 135C where it stabilized and slowly climbed.

It may be that all of the xylene isomers have an azeotrope with water and the azeotrope tables I use do not list it..
[Edited on 21-11-2016 by Meltonium]

[Edited on 21-11-2016 by Meltonium]