I have lots of p-xylene and I would like to make 1,4-benzenedicarboxylic acid from it. Obviously it has to be oxidized. I have been looking for a
procedure to this synthesis with no luck. I suppose that the reaction time is somewhat long and the product will be a mixture off all kinds of
isomers.Sauron - 7-4-2007 at 02:13
The acid is I believe called terephthalic acid, and as long as you art with pure p-xylene rather than mixed xylenes, you will get only the target
di-acid, the only possible contaminants being unreacted xylene (liquid) and 4-methylbenzoic acid. (aka p-toluic acid.)
The classical oxidizer for this is sodium dichromate/conc H2SO4. See Vogel, p.760 (which is p.785 of the DjVu file to be found in forum library, you
need a DjVu reader (available free from Lizard Tech) to read it.)
From 25 g (29 ml) p-xylene the yield given is 17 g purified terephthalic acid.
This involves 140 g of sodium dichromate 2-hydrate in 300 ml water to which 185 ml (340 g) conc H2SO4 is slowly introduced with stirring over 60-90
minutes with stirring.
The mixture is then refluxed 30 minutes.
As you might surmise this process generates a lot of aqueous Cr(VI) waste stream which is environmentally awful, corrosive, and carcinogenic. You
might want to look into alternative methods of oxidation, which might include KMnO4 aq, or nickel peroxide in 5% aq.NaOCl (aka, Clorox.)
Let me know if for some reason you can't get your hands on Vogel (Practical Organic Chemistry, 3rd ed.) and I will scan a printout of the page for you
and post it here as a pdf.
[Edited on 7-4-2007 by Sauron]not_important - 7-4-2007 at 02:48
If you want to convert the p-xylene, KMnO4 is the simplest route. If you are after terephthalic acid, then hydrolyzing PET bottles (soft drinks,
fruit juices) is the simplest and cheapest.
I am with my relatives this easter, but as soon as I get home I'll try the conduct the syntesis. I have both dichromate and conc. sulphuric acid but I
am not too fancy to waste that kind of ammounts into this. I think I'll start from micro-scale. The oxidation with potassium permanangate is also
interesting. I recieved ~50g of potassium permanangate from chemists last week, it had gone old so it didn't cost anything. I think that I'm going to
run a few tests with different ammounts permanganate and compare the yields (if any). The tereftlaic acid is not the most important here, I am mostly
after interesting synthesies that can be conducted with xylene.
[Edited on 7-4-2007 by kaviaari]Sauron - 7-4-2007 at 03:06
That's terephthalic acid.
The product from m-xylene would be isophthalic acid
And from o-xylene, phthalic acid itself.
You certainly should be able to find examples of permanganate oxidation of toluene to benzoic acid, or 3-methylpyridine to nicotinic acid (a real
classic, that one is in Org.Syn for sure.) So same procedure could be used for p-xylene, just increase the permanganate by 2X to take care of two
methyl groups instead of one.
Don't wear your best clothes, KMnO4 and MnO2 are messy.
[Edited on 7-4-2007 by Sauron]Klute - 7-4-2007 at 09:36
Oxidation of o-xylene to phtalic acid is a common school manipulation in my country, I did a few years ago, and they still do it now... Using
neutral KMnO4, you end up with butload of MnO2 though, which retains some product... Using acidic KMnO4 wouldn't be such a good idea as the product
would precipitate as form, and give a unstirrable mess, but neutral (basic) KMnO4 usually gives good yields... using of course 4x molar of oxidant.. I
can translate a protocole if you want, if I pass by my old school they will surely give me one. Heating at 80-90°C with vigorous stirring for an hour
of two IIRC, and filtering/washing the MNO2, neutralizing excess oxidant with bisulfite and acidifying IIRC, would seem logical.kaviaari - 7-4-2007 at 09:45
The total protocole would be nice if you can traslate it. I think that potassium permanganate will be my choise, foolin around with dichromates is way
more dangerous.Sauron - 7-4-2007 at 10:46
The freeing of the terephthalic acid from the MnO2 cake is just a function of repeated washing with good technique and finally pressing the cake with
a rubber dam. 3-5 washings ought to suffice. p-Xylene is cheap enough that a little mechanical loss of product ought not to phase you, unless you are
finicky about that sort of thing, but you are not doing this for publication.
Obviouly from a carboxyl group you can convert to ester, amide, acid halide, nitrile, amine, hydrazide, and so on. You can do that on eithger or both
groups on your d-acid, protecting the other while manipulating one. You could fill up a lab notebook or two making things out of the three xylenes (or
any one of them) and probably never exhaust the possibilities. (For example if you make an amide function and degrade it to amine you open up the
whole broad diazotization frontier which can get you a few dozen possibilities by itself.) So there's a few years of five-finger exercises there, but
it's more important just to understand them than to actually try to do them all, and move on to something more interesting.Klute - 7-4-2007 at 14:43
Sauron, would you know by any chance if seperating o,m and p isomers of the dicarboxylic acid is easier than the xylene isomers?Sauron - 7-4-2007 at 15:07
The three di-acids are all solids AFAIK, I would expect them to be easily seperated by LC. The xylenes are usually fractionated.
As all three xylene isomers are cheaply available (I have a gallon of o-xylene sitting around) and not much more costly than mixed xylened, I fail to
see why this would be necessary. In practice phthalic anhydride is most common, isophthalic acid least common, terephthalic acid of importance mostly
in polymer industry for PET resins.Phthalic acid is quite a bit costlier than the anhydride.
Merck Index probably holds clues as to the seperations as would the other usual suspects K-O and Ullmann.Klute - 7-4-2007 at 15:36
I only have access to mixte isomers xylene, and carefull fractionation with a 300mm vigreux would apparently not be effecient, so I was told. I guess
if I really wanted to find some seperate isomers, i could, but it was more of an idea that passed by while reading your little paragraph on the
numerous transformations one could do, a little collection of different isomers/ compouds sounded like a nice way of passing a weekend Sauron - 7-4-2007 at 16:02
I think that would occupy a lot of weekends as we are talking hundreds of compounds.
But anyway I will have a look and tell you what I can find. I can tell you this: a 300mm Vigreaux is not a very efficient column. In general a
Vigreaux is okay for easy simple seperations like stripping solvent from higher boiling product, but even a 1 meter Vigreaux isn't many theoretical
plates, the HETP is rather a large number. I bet for xylenes fractionation you are looking at a 15-20 tp column, and a pretty high RR (reflux ratio)
of similar qty. Because I think the bp's are pretty damned close.
The point you are getting at though is, go ahead and oxidize the mixed xylenes and then seperate the three acids. So let's see if any solubility
differences suggest seperation by crystallization, or maybe derivitization might make sense. Maybe an ester that could be fractionated more easily,
then saponified? Maybe a glimmer of hope there.Klute - 7-4-2007 at 17:36
Yes, that was exactly what i was thinking about..
As for the coluum, yes a vigreux is only limited to components having at least 30-40°c difference in their bp.. I'm actually thinking about working
on building a simple reflux control valve, regulated by simple electronic device (I'm not much into electronics, but hell they can be handy...) which
would give much better results though longer distillation times... So that why fractionating the isomers wasn't a possibilty...
Without spreading too far, i guess one or two "simple" transformations of each isomers could be pretty interesting without needing intense
investisment.. there's so much I'd like do to..
I manadge to find my old protocole on o-xylene oxidation! Apparently we used NaMnO4.H2O instead of permanganate, strange I don't remember incountering
the sodium salt before, but... here goes: 5 i simplified the document as alot had to do with questions, and useless describtion of the reflux setup
etc..)
Protocole: Preparation of benzene-1,2-dicarboxylic acid (ortho-phthalic acid)
This component is prepared by the addition of 1,2-dimethylbenzene (o-xylene) to an aqueuse solution of sodium permanganate in a RBF with magnetic
agitation and thermometer heated by a water bath.
- Insert 24.0g of NaMnO4.H2O (160g.mol-1) and about 150mL of water in the RBF.
-introduce 5mL of o-xylene in the addition funnel
-fill the water bath with hot water and heat with the hot plate set at maxium with vigorous stirring.
-when the temperature approches 50°c, drip in the o-xylene with vigorous stirring. The addition must take at least 5min.
(note: it is normal that all the NaMnO4 cristals haven't dissloved yet (me: IIRC, KMnO4 is even less soluble, so adding alittle more water could be a
good idea, even if it isn't very problematic))
-Once the addition is complete, let the reaction proceed for 45min from the point where the temperature reaches 70°c. keep the hot water bath
boiling. (me: during those manipulations, we often had only 4h to complete the reaction, the recrystallization, caracterisation and hand in the
report... So 45min is quite short, you could leave it for an hour or more at least IMHO)
-Whuile maintaning the agitation, cool the flask with a ice cold water bath.
-redraw a drop of the reaction fluid and deposit it on a piece of filter paper and look if the aqueus ring is pink or transparent (to see if any
permanganate remains)( check the pH after this, somethimes if you add alot of sulfite the pH becomes neutral or slightly acidic, keep it basic by
adding alittle base if need be)
-if any oxidant is left, add a spatula of sodium bisulfite with agitation to the RBF, repeat the test and the addition until no more oxidant remains.
-introduce 5g of Celite (for helping filteration) with stirring for a minute or two.
-filter the flask contents under vacuum, wash the solids with about 30mL of water while breaking the vacuum. (Could use alittle more or do another
washing, but if you have easy access to the isomer, don't bother as Sauron mentionned)
-Cool the filtrate in a beaker with an ice cold water bath, and while stirring with a glass rod introduce small amounts of conc. HCl until pH <1.
Stir for a few minutes with the glass rod until no more precipitate appears.
-Filter the white precipitate under vacuum and wash it with a little amount of ice cold water (20mL) while breaking the vacuum. Dry the solid on the
buchner under vacuum for a few minutes.
-dry the solids in a 85°c oven (dunno the specific english name for those "ovens" used in labs)
Caracterisation:
Taking the mp of phthalic acid isn't easy, so a little amount of the product is converted to the corresponding anhydride.
Place 0.5g of product in a small glass container with a "watch glass" on the top covered with a little cold water. Heat strongly on the hot plate and
observe. (I don'tr emember this part, I guess the anhydride appears on the watch glass (if that word exist in english) from sublimation..
Physical data:
o-xylene (1,2-dimethylbenzene) mw: 106g/mol solubility in water : none density: 0.878
phthalic acid (benzene-1,2-dicarboxylic acid) mw: 166g.mol solubilty in water: at 85°C weak, at 20 very weak pKa1=3.0 pKa2=5.3
sodium permanganate (monohydrate) mw: 160g/mol solubilty in water: at 85°c great at 20°C weak
manganese dioxyde mw: 87g.mol solubilty in water: very weak.
There you are, the whole thing needs a little twisting, as a school protocole doesn't really aim for maximum yields, and has only 4h to complete
everything...Sauron - 7-4-2007 at 22:20
I am sure NaMnO4 must be around, but in 40+ years as a chemist I have never seen it. KMnO4 is common as dirt. The sodium salt would require less by
weight. Which one is more expensive?kaviaari - 7-4-2007 at 22:44
I read that sodium permanganate can be replaced by potassium permanganate easily, so I'm going with that. At the moment I don't have sodium bisulfite,
can it be replaced by something more "common". Of course if it is a must it can be made from sodium carbonate and sulphur dioxide.
[Edited on 8-4-2007 by kaviaari]not_important - 7-4-2007 at 22:52
Ah, Sauron, perhaps you were looking on too small a scale
I did not say it did not exist. I said I have never seen it, and I meant, in a lab or in a stockroom for a lab or department.
I am quite sure it is in catalogs.
It is not very exotic after all.
I would guess it might be neglected because it may be hygroscopic, same reason the K nitrates and chlorates are often preferred.
---------------------
@klute, it appears that the p-bromophenacyl ester and the p-nitrobenzyl ester and the p-phenyl phenacyl esters have sharp well defined mps that are
well away from each other, and so I suspect that under vacuum they will be volatile and so a mixture could readily be seperated by fractionation with
a decent column.
See Vogel p.804 (table of derivatives of aromatic carboxylic acids.
That is page number of the DjVu document, logical page number not actual page number of print doc.
Use Navigate_GoTo Pagekaviaari - 7-4-2007 at 23:56
In the synthesis of adipic acid they use methanol to remove the excess of permanganate. Is it safe to replace the bisulfite with methanol, what about
sodium hydroxide?not_important - 7-4-2007 at 23:57
Quote:
Originally posted by Sauron
I would guess it might be neglected because it may be hygroscopic, same reason the K nitrates and chlorates are often preferred.
It is, usually delievered as solution in large amounts. I've seen in a stockroom once, but I've seen tankcars of it a dozen times or more.Sauron - 8-4-2007 at 03:42
NaOH would give you the Na salt of the di-acid, I'd expect.kaviaari - 8-4-2007 at 06:17
I started with 10g of potassium permanganate in about 83ml of water. At the temperature of 50°C I slowly dropped in 3ml p-xylene. Foolin around with
permanganate is quite messy, and the flask has gone quite dirty from dioxide. Well, luckily it is quite easy to clean up.
Sorry about the picture, it's not so good.
[Edited on 8-4-2007 by kaviaari]
Sauron - 8-4-2007 at 06:53
@klute:
The p-bromophenacyl monoesters have mp's as follows
phthalic 153 C
isophthalic 179 C
terephthalic 225 C
The p-nitrobenzyl monoesters are even better differentiated:
phthalic 155 C
isophthalic 203 C
terephthalic 264 C
I would favor the latter myself, as you can seperate thse even with a not so great column and a not so high RR.
p-nitrobenzyl alcohol is not hard to buy or make. So, assuming you have a good vacuum pump so you can volatilize the esters, there's your preparative
route from mixed xylenes via oxidation, to all three diacids. Just monoesterify the lot, then fractionate, then saponify the fractions and voila.
Oh yes, by the way, the acids themselves: (mp's)
phthalic c.208 C dec
isophthalic 340 C (Vogel) 345-348 C sublimes (Merck)
terephthalic 300 C sublimes
Too bad about phthalic acid, else there would almost be a practical seperation method by sublimation, given a largish sublimation apparatus.
Commercially available ones can handle about 50-100 g per batch.
I have dug up some articles and patents and will report on them as soon as I can review the contents.
So clearly fractionating the acids proper is not an option.
[Edited on 8-4-2007 by Sauron]
[Edited on 8-4-2007 by Sauron]Klute - 8-4-2007 at 08:13
Thanks alot sauron! I don't think i could nitrate benzyl alcohol, as that would surely oxidize the hydroxyl group... How about monoesterification,
followed by nitration? I'm not sure the ester would supporte that... I'll have a look, perhaps there's an easy way of protecting the benzylic -OH...
It's a pity phthalic acid decomposes, it would have been possible to seperate that from the other two...
I guess a very tough vacuum is needed, I'm going to unfortunaly wait to get the mecanical pump, as I can only get to 20mmHg for now... But that
leaves me time to work one the nitrobenzylalcohol and the mono-esters...
Thanks again Sauron!
BTW, NaMnO4 and KMnO4 are perfectly subsitutable IMHO. Just a question of solubilty, but that's no problem.. Maybe FeSO4.7H2O can replace the
bisulfite, it's often used to reduce excess oxidant, and is easily made by boiling clean Fe wool in dilute (30-40%) H2SO4 , filtering and then gently
evaporating (not over 50° or you get the monohydrate which is much less soluble) half of the solution and cooling it in the fridge: beutifull green
crystals. You can also make it by immerging stell woll in a Cu2+ (CuSO4 for exemple) solution until it doesn't get covered in Cu metal anymore, than
just gently evaporate half the solution like above... Need to check if it works for permangante, but i can't think of a reason it shouldn't...
EDIT: maybe a sort of fractional melting in a test tube heated in a fix oil bath at 160-170°c could seperate a small amount of the o-isomer, though
it would be imprecise, or only the terephthalic ester (leaving it solid) of course with certain amount of the other isomers...
[Edited on 8-4-2007 by Klute]Sauron - 8-4-2007 at 08:25
Bear in mind that if the particular esters aren't volatile enough then it would not be too hard to identify esters that are much more volatile, I
simply proposed these because the mp data was at hand. A few minutes with a nomograph will probably get you the bp at any given vacuum, if you can
find the bp at 760 mm. My first source (Vogel) did not give bp's but other reference works might.
So I would not worry too much about the nitration just yet.
Here's an interesting 1921 JACS article describing the KMnO4 oxidation of a fraction obtained from mixed xylenes using a 150 cm column. The resulting
acids (isophthalic and terephthalic) were mostly free of phthalic acid which means that the fractionation was effective at eliminating o-xylene. The
acids were isolated and purified first as the barium salts and then as dimethyl esters.
Hang on because I am moving on to a much more recent and interesting article now.
[Edited on 8-4-2007 by Sauron]
Attachment: ja01441a020[1].pdf (320kB) This file has been downloaded 837 times
not_important - 8-4-2007 at 08:58
Quote:
Originally posted by Klute
Sauron, would you know by any chance if seperating o,m and p isomers of the dicarboxylic acid is easier than the xylene isomers?
Separation by solubilities would seem to have some possibilities, as they are soluble o > m > p with rations of one to two orders of magnitude
difference. If you know the approximate amount of the isomers, adding enough NaOH or Na2CO3 to form the mono-sodium salt of the ortho isomer would be
very effective as it's roughly 4 orders of magnitude more more soluble than the m and p acid; heat the mix in sufficient water to dissolve the ortho
monosalt, and the calculated amount of base, let cool, and filter off the remaining m and p acids.
Heat based methods are another. Heating to dehydrate the ortho isomer to the anhydride and subliming that off could be practical depending on the
sublimation rate below 250 C.
In the Henkel process the potassium salts of benzoic or phthalic acids are rearranged to terephthalate under the influence of a Cd or Zn salt as
catalyst; the reaction is driven by the crystallising out of the higher melting and less soluble terephthalate.
And now I see that the PDF Sauron recently left also uses differential solubility.
[Edited on 8-4-2007 by not_important]kaviaari - 8-4-2007 at 09:04
The mixture has been refluxing for an hour and a half now. There is no potassium permanganate in the mixture whatsoever. I don't have acces to vacuum
source so I filter the MnO2 off with glass funnel and cotton plug. There is a somewhat large ammount of water in the mixture, so I think that a part
of it oght to be evaporated before adding the hydrochloric acid.
[Edited on 8-4-2007 by kaviaari]Sauron - 8-4-2007 at 09:28
US Patent 2 853 054 (Brill, to Olin Matheson) describes oxidation of various compounds to arylcarboxylic acid using cobaltic acetate and O2 at 1 atm
and relatively mild conditions. Substartes possible include all three xylene isomers as well as mixed m- and p-xylenes, as well as others. What is
particularly useful is that the inventor describes many details of reaction variables when conducted on bench scale in an article in Ind.Engin.Chem.
which I have bundled with the patent. A ketone is used along with the catalyst and USP O2 but this is merely MEK which should present no difficulties.
The reaction time and temperature are very similar to those cited in the 1921 JACS article for permanganate. While it does require oxygen, the
procedure at least eschews the mess of MnO2, affords 93% yield and signals end of its induction period with a convenient color change (purple to dark
green). I think it deserves close scrutiny
[Edited on 9-4-2007 by Sauron]
Attachment: 2853514[1].zip (921kB) This file has been downloaded 595 times
Klute - 8-4-2007 at 19:40
Nice info...
Good luck filtering the MnO2 without any vacuum! Do at least a couple of washings in that case or quite alot of product will saty trapped in the
solid...Sauron - 8-4-2007 at 20:04
Do yourself a big favor and get either a water aspirator or a hand operated Nalgene vac pump.
The former (also Nalgene) works best with cold water, which in Finland I bet is in good supply from the tap.
The latter, looks like a hand exercizer, come without or without gauge (without is cheaper) and works well. You will also have strong hands.
Then get yourself a few fritted-glass Buchner funnels (not Coors perf plate) in various sizes and porosities. And rubber adapters so you get a decent
seal between funnel and flask, connect vac hose between pump and flask and you will REALLY save yourself a lot of aggravation and wasted time waiting
for gravity filtrations.
------------
The dimethyl and diethyl esters of the three di-acids have bp's very close to each other and so are no seperative help.
The p-nitrobenzyl ester still appears to be best bet.
This ester is prepared from the sodium salt of the acid and p-nitrobenzyl bromide not the alcohol.
The corresponding p-bromophenacyl ester is prepared from the sodium salt of the acid and p-bromophenacyl bromide, aka 4-bromo-bromoacetophenone, a
very very nasty lachrymator. So I would avoid this one and make the p-nitrobenzyl ester instead. While it is also lachrymatory it is not in the
military/police tear gas category, while phenacyl bride is a famous one of those and its p-bromo derivative would give no relief.
[Edited on 9-4-2007 by Sauron]
[Edited on 10-4-2007 by Sauron]kaviaari - 8-4-2007 at 21:52
Quote:
Originally posted by Sauron
Do yourself a big favor and get either a water aspirator or a hand operated Nalgene vac pump.
Overnight about 50ml of solution has passed the filter, this is really going to take some time.
I have a filtering flask, rubber adaptor and a buchner funnel but no vacuum source. I have been dreaming of a water jet pump, but I just can't find a
supplier for those. Mail ordering would be one option, but everything is somewhat expensive here.
I think that I could borrow an old jet pump from school, it has not been used for who knows how long.
[Edited on 9-4-2007 by kaviaari]Sauron - 8-4-2007 at 22:11
Org.Syn. has a prep for p-nitrobenzyl bromide on a 250-280 g purified product scale starting from p-nitrotoluene (which is easy) and using Br2 as
brominating agent. File is attached.
p-nitrobenzyl bromide is expensive to buy at $1-2$2 a gram so I'd just make it.
Org.Syn also has a procedure to make p-nitrobenzaldehyde from same starting material, and obviously you could reduce that back to p-nitrobenzyl
alcohol with a judicious choice of red.agent. You could also do a Cannizaro rxn and obtain both the alcohol and the p-nitrobenzoic acid.
Then brominate the alcohol with HBr conc. or else NaBr/H2SO4, this alternate route is more work but may be useful if you can't obtain or make bromine.
Incidentally, if you get bored starting with xylenes, you can make terephtalic acid quite easily from p-cymene, 4-methyl-isopropylbenzene. This is a
constituent of oils of cumin and thyme.
Might spark your interest in monoterpenes and natural products chemistry.
The xylenes and the isomeric phthalic acids are boring industrial chemistry by and large, mostly plastics constituents.
[Edited on 9-4-2007 by Sauron]
Attachment: CV2P0443[1].pdf (152kB) This file has been downloaded 786 times
Klute - 9-4-2007 at 14:14
Thanks for all this Sauron! Indeed the synth from nitro toluene is accesable, and the ester formation give an interesting little project all by it's
self.. I've only rarely stumbled on ester synthesis from acid salts and halogenoderivatives, so thta kind of a new feild to play in..
Can this method be applied to generally all esters formation?
I'm wondering is it could be possible to brominate benzylalcohol and then nitrate it, altough benzylbromide is as bad or worse then benzylchloride to
deal with IIRC. ANd I guess I won't be messing around with the bromo-phenacyl esters, we are getting enough lachrymatory gas as it is in my country :S
so no need to go looking for it
BTW penacyl bride a typo i guess, at first i thought is was some sort of trivial name Funny one if it was
Yes, I'm already having a little interest in compouds in natural oils, but generally the price of these oils doesn't allow much different reactions..
Also a few of them are quite a mess to fractionnate, but at least it smells NICE in the lab (not the residu from vacuum fractionnating cinnamon oil though!) The smell issue might be a major reason why i prefer working on ketones,
esters and aldehydes than , for example, amines
Anyway, thanks for your efforts to sparkle my interest as you say, it is very appreciated. You would do well as chemistry teacher of some sort, take
that as a compliment!Sauron - 10-4-2007 at 13:13
Yes I meant phenacyl bromide, best avoided. Yes you can brominate benzyl alcohol with HBr or NaBr/H2SO4 quite easily. BzBr is a little nasty but,
life's a bitch. IMO that procedure is probably a lot lkess nasty than bromination of toluene with the element.
Nitrating the benzyl bromide you will get both isomers and will need to seperate them but doubtless that is in the literature somewhere for guidance.
In fact the Org Syn procedure has a ref. to that in Annalen, so as long as your German is up to it (archaic chemical german at that) you should be
fine.
I am quite sure these are the mono-esters we are talking about as I suspect the di-esters MW would be high enough to push up the bp too much. It would
be nice if we could find the bp's of these esters (we have the mp's) in the lit si we can verify that they will in fact distill, and not decompose,
and be far enough apart. Because if not then all this is for nothing.
