Sciencemadness Discussion Board

Synthesis of terephthalic acid from a PET water bottle

palico - 2-2-2024 at 12:20

Hello people,

I am back with another episode of chemical recycling, and our victim today will be again a waste water bottle. Let's find out!

The last episode was about the preparation of benzene by pyrolysis of waste PET water bottles, this time instead, from the same waste material we synthesize terephthalic acid.

The reaction involved is simple acid hydrolysis of PET, polyethylene terephthalate, into its precursors acid and glycol. The latter will not be recovered though.

reaction.jpg - 164kB

Procedure

A water bottle is removed from cap, seal, label, top, and bottom. The body is cut into pieces; 20 grams of that are introduced into a large beaker. Apart, 20 ml of distilled water is brought to 150 ml with concentrated sulfuric acid. The solution gets very hot. While still hot, the acid solution is poured into the PET pieces and stirred slowly. PET flakes soften, then melt and dissolve completely. Once done, the reaction mixture is kept under stirring for an additional 1.5 hours, then an equal amount of water is added - 150 ml: a grey suspension forms, mixture heats up. Once cool, an additional 1 liter of water is added, and the suspension gravity filtered. The residue is oven-dried, then dissolved with 250 ml of aqueous KOH. The solution is gravity filtered, the filtrate is slowly added of 100 ml HCl 37%: a white suspension forms. That is stirred for a while, then vacuum filtered, washed with methanol, and oven-dried.

5.83 g of white solid is left, 29.4% of the initial PET mass.
m.p: > 300 Celsius

IMG_20220827_154852.jpg - 875kB

Discussion

I think this experiment is wonderful, and considered one of my best chemical recycling examples. From total trash, a useful chemical can be obtained, which in turn can be converted to something else. The yield can be improved, I think, by exploring different conditions, as well as cost using NaOH rather than KOH, a minimum amount of HCl, ethanol in place of methanol.

I could measure the proton NMR spectrum of product in d-DMSO and it looked very pure !

As usual, I link my YT video, both preparation and analysis.

Thanks for attention

See you next time, with the esterification of that acid,

palico

[Edited on 2-2-2024 by palico]

Fery - 2-2-2024 at 21:13

Well done! Nice experiment! Yes, much cleaner than pyrolysis. Costs could be perhaps reduced by avoiding H2SO4 and hydrolyzing directly in sol. of NaOH? I wonder whether less of hydroxide as well acid cannot be sufficient for only 20 g of PET?

palico - 3-2-2024 at 11:32

Quote: Originally posted by Fery  
Well done! Nice experiment! Yes, much cleaner than pyrolysis. Costs could be perhaps reduced by avoiding H2SO4 and hydrolyzing directly in sol. of NaOH? I wonder whether less of hydroxide as well acid cannot be sufficient for only 20 g of PET?


PET hydrolysis in NaOH, requires long reaction time and high temperature.
Yes, use a minimum amount of HCl and NaOH is better to reduce production cost.

kmno4 - 3-2-2024 at 14:58

It is very unusual way of hydrolysing of PET.
A method for very rich and very mad madscientists, I would say.
It apparently works, but it is hard to call it useful, when someone converts 10 dollars into 10 cents.

PET is easily hydrolysed by > 20% NaOH, r.t. is good, heating speeds up the reaction, but it is optional. Disodium TPA salt is obtained in solution. Next, TPA can be precipitated with some acid, for example citric one, to reduce costs. NaOH method is described in many places, also at SM, and I am not going to bother readers with commonly known things.

Texium - 4-2-2024 at 08:00

There’s nothing wrong with demonstrating alternative methods. It may be easy to forget that outside of the EU, conc. sulfuric acid is readily available and cheap in most of the world.

palico - 4-2-2024 at 14:13

Quote: Originally posted by kmno4  
It is very unusual way of hydrolysing of PET.
A method for very rich and very mad madscientists, I would say.
It apparently works, but it is hard to call it useful, when someone converts 10 dollars into 10 cents.

PET is easily hydrolysed by > 20% NaOH, r.t. is good, heating speeds up the reaction, but it is optional. Disodium TPA salt is obtained in solution. Next, TPA can be precipitated with some acid, for example citric one, to reduce costs. NaOH method is described in many places, also at SM, and I am not going to bother readers with commonly known things.


Have you ever tried such reaction for real ? Are you sure you got TPA ? Because such procedure does not eliminate oligomers, I think. Mine, as you can see from proton NMR, give very pure TPA.

kmno4 - 4-2-2024 at 22:37

Quote: Originally posted by Texium  
There’s nothing wrong with demonstrating alternative methods. It may be easy to forget that outside of the EU, conc. sulfuric acid is readily available and cheap in most of the world.

As usually, you have something hopeless to say. You start getting on my nerves.

Texium - 5-2-2024 at 06:16

Quote: Originally posted by kmno4  
Quote: Originally posted by Texium  
There’s nothing wrong with demonstrating alternative methods. It may be easy to forget that outside of the EU, conc. sulfuric acid is readily available and cheap in most of the world.

As usually, you have something hopeless to say. You start getting on my nerves.
Oh really? Good. At least this time you actually say it instead of just deleting the post I called out like you usually do.

Boffis - 6-2-2024 at 10:42

Now now ladies, please put those handbags away!

This is an interesting method for the hydrolysis of PET resin but as a preparatory method it seem to have limited use. As kmno4 says the large amounts of chemicals, no matter how cheap and available, for a 34% yield based on the assumption that the PET is 100% pure, is not great. Does the original poster have any idea where the other 66% of the terephthalic acid has gone?

The alkali hydrolysis method carried out carefully yields 75-90% and if the liquor is recycled this can be raised to 95% easily. Only NaOH, water, dilute HCl and ethylene glycol are required and the latter is not essential but does reduce the reaction time by allowing higher temperatures.

bnull - 6-2-2024 at 12:07

Quote: Originally posted by Boffis  
Does the original poster have any idea where the other 66% of the terephthalic acid has gone?


In the discarded solution. If the PET didn't hydrolise completely, there may have been oligomers terminated by glycol. Ethanol reacts with sulfuric acid to form ethyl sulfate as long as the temperature is below 140°C; I suppose the same happened to the oligomers, which reacted with the acid (because there was too much of it anyway), forming water-soluble compounds.

For each terephthalic acid that leaves, there are at most two glycol terminations free to be attacked. With two terminations forming sulfate, the oligomer is much more soluble. After the addition of water, the terephthalic acid precipitated, but not the oligomer sulfate.

@palico, did you keep the solution? If you kept it, you could try to precipitate the partially reacted stuff (supposing my idea makes sense, of course) by using NaOH to neutralize the acid. Maybe it works.

Or maybe I'm dead wrong.

palico - 7-2-2024 at 10:08

@Boffis: I do not know what is your 34% yield. We do not know yield because the molecular weight of PET is not known, and of course it is a mixture. 5.84 g of TPA are the 29.4% of 20 g of PET.

bnull: your recipe require heating for long time at high temperature, energy cost today a lot, and also glycol. The acid hydrolysis instead can be realized with household chemicals. Anyway, have ever measuread at least proton NMR of your TPA got from basic glycol procedure ?

Bedlasky - 7-2-2024 at 11:08

Very interesting procedure! I add it to list of my future experiments. I appreciate your detailed report, unlike kmno4.

Quote: Originally posted by palico  
@Boffis: I do not know what is your 34% yield. We do not know yield because the molecular weight of PET is not known, and of course it is a mixture. 5.84 g of TPA are the 29.4% of 20 g of PET.


Just consider molar mass of monomer. Monomer contains 166/210 = 79,05 % of terephtalic acid. That's 0,7905*20 = 15,81 g of terephtalic acid in 20 g of PET. You isolated 5,84 g, so 5,84/15,81 = 36,94 %.

bnull - 7-2-2024 at 11:46

Quote: Originally posted by palico  
bnull: your recipe require heating for long time at high temperature, energy cost today a lot, and also glycol.


My recipe? What recipe? It was @kmno4 who first suggested the alkaline hydrolysis, not me. For fleck's sake... I just offered an explanation as to why the yield was not even close to what one obtains from the NaOH procedure.

I'll ask again: did you discard the acidic solution? Would you care to verify, if you still have it, what happens when you neutralise it with a hydroxide? I'm very interested to know if something precipitates out of the solution or the solution gets cloudy or a gel forms or whatever the flaming hell happens, if something at all. Yes, I'm that curious.

If, and only if, a suspicious crud sinks to the bottom or rises to the top (I don't know what is the density of all the oligomers, you see), there will be the rest of the TPA, still attached to at least two glycol terminations.

And yes, I've read some very interesting papers about hydrolysis of PET in concentrated H2SO4, all of them involving heating the mixture. The yield was above 95%.

[Edited on 7-2-2024 by bnull]

And also yes, your idea is quite interesting. I should have mentioned it before. I appreciate alternative ways to perform syntheses, or decompositions as it seems to be the present instance.

[Edited on 7-2-2024 by bnull]

palico - 8-2-2024 at 07:12

@Bedlasky: ah, I see yield is more than conversion, good. Anyway, I do not believe the basic procedure got 95% yield, I want to see proton NMR first.

@bnull: the mixture is heated in my experiment. Mixing water and sulfuric acid makes the solution get very hot, and I poured while it still was. I think longer reaction time give higher yield.

bnull - 8-2-2024 at 09:02

Quote: Originally posted by palico  
Mixing water and sulfuric acid makes the solution get very hot, and I poured while it still was.


What was the temperature?

Quote: Originally posted by palico  
the mixture is heated in my experiment. [...] I think longer reaction time give higher yield.


With continous heating, I suppose.

Boffis - 8-2-2024 at 11:57

Quote: Originally posted by Bedlasky  
Very interesting procedure! I add it to list of my future experiments. I appreciate your detailed report, unlike kmno4.

Quote: Originally posted by palico  
@Boffis: I do not know what is your 34% yield. We do not know yield because the molecular weight of PET is not known, and of course it is a mixture. 5.84 g of TPA are the 29.4% of 20 g of PET.


Just consider molar mass of monomer. Monomer contains 166/210 = 79,05 % of terephtalic acid. That's 0,7905*20 = 15,81 g of terephtalic acid in 20 g of PET. You isolated 5,84 g, so 5,84/15,81 = 36,94 %.


Hi Bedlaski/palico. No! its not the monomer you have to consider, its the monomer less one mole of water (lost through the ester linkage). So the MW of the monomer is 210.18 but the MW of the simplest monomeric unit is only 192.17. so in the first case the theoretical yield is 15.81g of Terephthalic acid while in the second case it is 17.29g. thus palico's yield of 5.84g is 5.84/17.29 x 100 = 33.77% not 36%.

However, 2% difference hardly negates the validity of my question. And I suspect bnull's answer is close to the truth. Terephthalic acid is sparingly soluble even in fairly strong sulphuric acid so I suspect that your hydrolysis was incomplete.

kmno4 - 9-2-2024 at 10:13

It seems, that above insane procedure is taken from patent literature (US3952053 or US4355175), with totally changed amounts of reagents. The original proportions PET/H2SO4 from the patents are much more "user friendy". This is info is for lazy Americans from the most of the world, but there is no need to thank.
The NaOH method is still preferred, because less steps are needed, no need to use H2SO4/HCl/HNO3/H3PO4 at all, recovery of ethylene glycol is much simpler (if someone needs).

palico - 9-2-2024 at 14:24

@bnull: Yes, looks a contradiction, man I am not native english speaker. The r. mixture is naturally heated by the sulfuric acid solution. That's all, no additional hotplate heating applied. No, I did not measure temperature.

@Boffis: I continue to trust more conversion than yield in this case. Yes, conversion or yield, of course reaction did not go 100% completion. Next time, I will try longer reaction time.

@kmno4: I do not remember reference for this procedure. I did not use any nitric or phosphoric acid.

bnull - 10-2-2024 at 04:41

Before the this thread degenerates into a flame war (I hope not), I want to ask some questions about the procedure and maybe even suggest some ideas. Why? Because I'm f***ing curious.

@palico: First, I want to say that there was no contradiction. When I quoted you, "the mixture is heated in my experiment. [...] I think longer reaction time give higher yield," I suggested continuous heating. To rely solely in the heat of hydration (?) of the acid to develop the reaction was one of the problems that I saw in the procedure.

What was the concentration of the acid you used? From the papers I've read (no patents, by the way), the acid hydrolysis with H2SO4 is very sensitive to concentration, being fast when the concentration is above 80% and terribly slow when below 70%. Considering that you added water to the acid, your final concentration was about 0.88 of the original strength, and after the addition of water it could be too weak to completely hydrolise the PET in 1.5 hours.

The other question is, did you discard the solution? I'm still very interested in it because of the substances that may be there. One of them is BHET, or bis(2-hydroxy-ethyl) terephthalate, which is one of the oligomers I described in February 6th. You can react BHET with a diacid, like phthalic acid or even oxalic acid, to produce a polymer.

My suggestion is, use 20-25 mL of concentrated (83% or more) acid (no water added) for each 5 grams of PET with constant heating for 3-5 hours. With concentrated nitric acid, you'll have oxalic acid rather than ethylene glycol, a more useful product. Neutralise the acid with NaOH, and the rest you know better than I do.

palico - 10-2-2024 at 13:09

@bnull: Yes I discarded the solution. No glycol recovery. Oxidise it to oxalic acid is a good idea.

Sir_Gawain - 19-2-2024 at 17:10

Heating is absolutely necessary. I tried this procedure with 20 g of finely chopped pet plastic and ~100 ml of sulfuric acid. The sulfuric acid was hot from the water added, and most of the plastic dissolved into a clear, thick liquid. After letting it sit for a day, there was still some undissolved solids so I decided to heat it. After a few hours at ~60 C, the solution turned white with precipitate. I think this is terephthalic acid dropping out of solution due to the large amount of it. I then poured the slurry into about 500 ml of water and filtered out the terephthalic acid. There looks to be a lot of it, definitely more than 5 grams. Currently it’s drying, I’ll report back after I dissolve in sodium hydroxide solution and reprecipitate the terephthalic acid.

Also, while I was filtering it, there was a distinct smell of ethylene glycol.

[Edited on 2-20-2024 by Sir_Gawain]

Fery - 19-2-2024 at 21:51

Ethylene glycol is odorless. Why not hydrolyzing PET directly in solution of NaOH and avoiding wasting H2SO4?

Rainwater - 20-2-2024 at 02:02

Quote: Originally posted by Fery  
Ethylene glycol is odorless.
when distilling eg i have noticed a smell, unique and i can only describe it has "leaking heater core" smell without the metallic tinge. But my source was antifreeze and likly to contain impurities

[Edited on 20-2-2024 by Rainwater]

j_sum1 - 20-2-2024 at 03:56

I am currently processing about 60g of coke bottles in concenrated NaOH solution (approx 4x excess). It has been sitting for over a week and has had several hours of refluxing at about 110C during that time. A few chunks remain but it has mostly broken down.

I now have a white powder settled amongst the remaining fragments of unreacted PET. I am assuming this is the disodium salt which is undissolved due to my liquid volume being so small and hence the solution concentrated. I am yet to do proper testing on the material. But another week or two sitting in the beaker will not hurt it.

It is really nice to have an experiment that can just sit and do its thing while I have been caught up with work and family and other pressures.

No idea what I will do with the terephthalic acid once I have it. (The only use I know of is making PET and I don't think I need to do that.)

Texium - 20-2-2024 at 06:18

Quote: Originally posted by j_sum1  
No idea what I will do with the terephthalic acid once I have it. (The only use I know of is making PET and I don't think I need to do that.)
Unfortunately it’s not a very interesting molecule. You can decarboxylate it to benzene, but that’s more practical to do with sodium benzoate. You could convert it to the diacid chloride or the dimethyl ester, but these are just ways to turn a boring compound into slightly less boring but still not too useful compounds.

palico - 20-2-2024 at 07:38

@ Sir Gawain: Take in count this hydrolysis is very sensitive to acid concentration. Once I attempted hydrolysis in 50% sulfuric acid, heated to more than 100 Celsius overnight, got no hydrolysis at all.

@Rainwater: the smell can be glyoxal.

@ j_sum1: you can convert it to its diethyl ester. I made a post about.

@Texium: The diethyl ester can be used as high boiling solvent.

Boffis - 27-2-2024 at 07:54

The best way forward to interesting derivative from terephthalic acid is to nitrate it as per kmno4 posts in erlier threads on terephthalic acid from PET. From this can be prepared aminoterephthalic acid, the acetanidilide and then to a whole raft of compounds but this area of chemistry is not well explored.

kmno4 - 27-2-2024 at 10:03

Having some free time, I decided to try patent (US4355175) procedure, just for sheer fun.

It means 10g of PET (bottle, colorless) + 18g of conc. H2SO4 + 2g of H2O.The mixture soon becomes syrupy and pinkish. After initial heat effect (H2SO4+H2O), it was kept at ~30 C during few days. The mixture becomes more and more sticky, some white sediment is formed, surely TPA.The weak odour of the mixture reminds me experiments with H2SO4 and EtOH, most possibly some kind(s) of "glycolhydrogensulfate" and/or etheric products are formed. No 1,4-dioxane odour is detected at this stage and later. Visibly, some PET praticles dissolved.
The sticky mixture was transfered into beaker with water (heat is generated). Colorless, coloidal solution is formed with white sediment and unreacted PET. It was heated for some time, solution becomes transparent, some PET-gel precipitates (looks like ion-exchange resin), but not much. The solution was then filtered, filtrate was kept.
The sediment was washed with NaOH solution, boiled, filtered and TPA precipitated with saved earlier acidic filtrate. About 4g of TPA was obtained. The PET-based slightly brownish residue (~6g) was dried and reused for alkaline hydrolysis.

Alkaline hydrolysis.

13 g of PET (cut bottle, ~ 0,5 cm x 0,5 cm pieces) was placed in a screw-top glass jar, 10 g of NaOH (small granules) and 10 g of water are added. The jar (closed) was then heated on water bath (almost boiling) for several hours. Slowly, gradually, PET hydrolyses, white sediment is formed (TPA Na salt), the mixture becomes semi-solid mud. From time to time, the jar is tapped by hand to obtain more uniform mixture.
At the end of reaction (or your patience), when PET particles ar not visible, the jar is cooled, the mixure is dissolved in ~200 cm3 of water, heated, filtered, some acid is added (e.g. citric)... etc. When time of NaOH treatment is long enough, conversion of PET into TPA is practically 100%, but it also depends on input PET form*. The reaction can be easily scaled up, without any problems, time of NaOH treatment remains the same.

Some publications suggest the use of KOH, as better "solvent" for PET than NaOH - it was also tried.

11 g of PET + 10 g of KOH (commercial, ~90%) + 10 g of water and the rest as above. The reaction is much faster than with NaOH, the reaction mixture is also much less thick. It is caused partly by better TPA K salt solubility, but the main reason is "just" larger KOH activity. It is another proof, that conc. NaOH is not repacement for conc. KOH.
In one experiment, the end-white-mud was added to partly reacted PET+NaOH mixture. Effect was almost the same as in case of PET+KOH only reaction: PET particles disinegrate much faster in NaOH+KOH mixture than in NaOH only... interesting.

I am going to make some titration of prepared TPA, but my standard NaOH solution is almost out. I have to buy or prepare it.

* a radical improvement in shortening of time of PET hydrolysis is using "powdered" PET. It can be done with
stright saw-blades (several such blades joined in prarallel) and PET plactic. The best for this purpose are the thickest parts of PET bottles : tops and bottoms. Obtaining this "powder" in such way, manually, is very boring job, possibly it can be done with some mechanical tools. However, time of reaction with PET-powder is about 1/10 of reaction time with PET-cut material, also in case of less concentrated alkali.

Additional notes.

1)When at some stage, water solution of TPA salt is colored, the coloration possibly can be removed by small amout of sodium percarbonate on cold. It is important, becuse in hot solutions percarbonate decomposes itsef, without acting on colored impurities.

2)In each case of akaline hydrolysis, weak alkoholic odour of glycol is sensible. With KOH, small brownish layer is present on top of post-reaction mud. But its recovery is much more time consuming than TPA.

3) Experiments with PET + molten citric acid and PET + molten urea were conducted. Unfortunately, PET remains unchanged, at least during few minutes and near melting point.

Texium - 27-2-2024 at 11:23

What were your yields of TPA from the NaOH and KOH hydrolysis runs?

palico - 1-3-2024 at 23:37

@kmno4: have you confirmed your product be TPA ?

kmno4 - 2-3-2024 at 08:30

Having again some free time, another alkaline hydrolysis experiment was performed to test slightly lower hydroxide loading.

This time, doomed to destruction was 1 liter PET bottle (=PB), containing earlier canola oil. The PB was washed inside with detergent several times, but very weak odour of the oil was still present after drying. However, the PB was clean, with traces of some glue (from label). The top-part and the bottom-part of the PB were "powdered" as given above, the rest was cut ( shears for steel sheets are very good for this purpose). In this way, 23 g of pre-depolimerized PET was obtained.
A jar, ~160 cm3 (as above) was taken, the PET was paced inside, 8 g of KOH (old) and 8 g of NaOH (very old) were added to the jar, mixed and 16 g of water was added. Nothing interesting happens, some heat is generated, that is all. The bulk volume of the starting mixture does not exceed 50% vollume of the jar. The jar is then immerced in water bath (as above). During 2-3 hours of heating, whitish mud is formed, most of PET disappers. Aftrer this time, inspection reveals unreacted PET particles inside the mud. Additional ~5 hours heating was applied, but still some PET was present. The jar was cooled, the mud was transferred into some beaker with water (200-300 cm3). It was heated until all salts are dissolved. Some brownish scum was present, some PET particles separates, the liquid is clean and slightly yellowish. It was then filtered. The amount of brownish scum was very small, the PET was separated and dried. It looks like ~5 g, but in reality it was 1,7 g. To the clean water filtrate, some acid was added to precipitae TPA. White suspension is formed. It was heated up to boiling (shortly) and filtered.
The filtered, washed many times with warm water and dried crude (but white) TPA weighs 17,5 g.
ps. the mass of the jar before and after heating was exactly the same

PS.
Standard NaOH solution was prepared and a sample TPA from previous runs (combined) was titrated. Assuming that it is dibasic acid, calculated molar mass is 165,3.

Titration of a sample of crude TPA from the last run (17,5 g ) gave M=165,7. It means that some mysterious "oligomers" exist only in the doubters' minds.

[Edited on 3-3-2024 by kmno4]

bnull - 2-3-2024 at 10:39

I want to add my five kopeks to the discussion. I read some weeks ago a paper (Towards closed-loop recycling of multilayer and coloured PET plastic waste by alkaline hydrolysis) that proposed a procedure with conditions so mild that I was tempted enough to tryNote 1. It was February 7th, I believe (I don't keep track of trial runs).

First of all, I don't have a NMR spectroscope, so don't expect to see fancy graphs with well formed TPA peaks. I'll take some pictures of the products in two weeks (give or take a week) and post here.

The procedure was the following. According to the authors, the optimal conditions for PET hydrolysis are met by a solution of ethanol in water (60 vol% EtOH, 40 vol% water) with 5% in weight of NaOH. The temperature should be kept at 80°C. For 10 mL of solutionNote 2, I first dissolved 0.5 g of NaOH in 1 mL of water and completed to 4 mL. To that were added 6 mL of ethanol azeotrope (96.5%). The solution was moved to a glass bulb heated in a water bath (a paint can with water at 90°C) until equilibrium.

A strip (about 3 g) of PET, kindly donated by a water bottle and previuosly cut into pieces of 5 mm or less, was added at this moment. I waited about 2 minutes and removed the flask from the bath. To my surprise, the pieces of plastic showed signs of attack. They presented fine parallel grooves, as if they had been sandpapered.

It was at this moment that the ghost of Gergel intervened. While adjusting the height of the support, my hand slipped and the flask tumbled, spilling about half of the solution. I gave up, transferred the remaining solution and the pieces of PET to a small polypropylene vial, which is in my drawer, and cleaned the place.

Two or three days later, some of the plastic fragments had a white mass growing at the corners: sodium terephthalateNote 3 had formed. Yesterday, I looked again. The growth slowed considerably, and small crystals appeared in the bottom of the vial. They are similar in appearance to fine table salt or boric acid. As the reaction is still going on, albeit very slowly, I'll leave the vial as it is, except for a piece of etched PET and a grain from the white mass.

Note 1: My lab has been languishing in neglect for some years since I begun studying Physics. Few reagents left and very crude equipment. My own fault, you see. Nevermind.

Note 2: The quantities are slightly off because the NaOH absorbed some water from the air (rainy day) and the volume of alcohol-water solutions is always less than it should be. Let's ignore the small difference (0.2 mL perhaps) and say the volume was 10 mL.

Note 3: It may be either disodium terephthalate or a sodium salt of some oligomer. If we may judge from the solubility in water, it is the disodium TPA salt. And the oligomer salt (sodium-terephthalic acid-glycol-terephthalic acid-sodium, for example) would have been attacked by the dissolved hydroxide anyway. I'll pyrolise a sample to see what happens.

[Edited on 2-3-2024 by bnull]

kmno4 - 7-3-2024 at 14:19

... and the last PET destruction experiment was performed, to test another idea.
Intensively green-colored PB, with rather thick walls, was cut and powdered (as above). 25,5 g of obtained PET crumbs were placed in the jar (as above) with 8,2 g of KOH + 8,2 g of NaOH + 17 g of water. The jar was covered with watch glass and placed in microwave oven. The power was set at 90 W and 2-4 minutes. After several heating cycles, white-blue mud is formed, bulk volume of the input decreases. From time to time, the mass of the jar is checked, if the mass loss is greater than 2-3 g (evaporation), appropriate amount of water is added to keep total mass constant (more or less). Also from time to time, the content of the jar is mixing with some rod, to keep the mass as uniform and compact as possible. Finally, all PET particles are suspended in the mud. It is heated for some time longer, as hydroxides are consumed, the reaction becomes slower and slower. I had no time for it, the colored mud was transferred to the beaker with water and heated (as bove). Obtained liqiud was intensively green-blue colored, but inspection reveals that the dye forms "clouds". The liquid was filtered, giving clear yellowish solution, deep blue sediment (looks like Prussian blue) in small yiled and ~5.7 g of green PET chips. After earlier experiments, 4,2 g of colored TPA was left and it was added to the solution. The solution was processed as bove and gave 19,6 g of crude TPA. I have impression that its colour is not so purely white as TPA obtained from colorless PB. But it is white :)
Anyway, MW heating is interesting modification and as above, it seems to be easily scalable procedure. It this way, my boring alkali-thread is finished.
Good fight - good night !

RU_KLO - 8-3-2024 at 11:34

just more information

"Alkaline Hydrolysis of Polyethylene Terephthalate at Lower Reaction Temperature"
Masakazu YAMASHITA* and Hideyuki MUKAI*

https://doshisha.repo.nii.ac.jp/record/21417/files/023052020...

Dr.Bob - 9-3-2024 at 17:25

Yiu can also directly make dimethyl TPA by heating the polyester in NaOMe in methanol. That is one of the ways it is done commercially, according to a CEN article. That allows the ester to be distilled for higher purity. The EG can also be recovered from that reaction somehow, but not sure how. They claim that gives good yields and a way to very pure diester.

kmno4 - 10-3-2024 at 15:41

A small addendum.

Combined crude TPA from several alkaline-hydrolysis (72 g) was additionally purified by dissolving in NaOH sol., filtered and precipitated by acid treatment. Amount of dried TPA was 70,8 g, with mechanical losses ~0,5-1 g.
It means that the crude TPA is ~99%, as it is suggested by titration. If all operations are carefully performed, this addidional purification is - in most cases - pointless.

my results

dicyanin - 24-3-2024 at 12:22

Experiment 1
In a 500 ml screw-top glass jar were placed 25 g PET[1],[2], 25 g KOH (anhydrous flakes) and 30 ml distilled water, and the mixture was heated on a boiling water bath for several hours. After standing overnight, 10 g NaOH and 10 ml water was added and the mixture again heated on a boiling water bath for 5 hours by which most of the PET was consumed and a white precipitate was noted. 250 ml near-boiling water was added and the mixture filtrated twice through 2 coffee-filters.

The clear filtrate smelled of glycol. It was then slowly acidified with azeotropic aqueous HCl. First a white precipitate was noted, which redissolved upon stirring, the pH was checked with indicator paper and the mixture was still alkaline. Upon further addition of aq. HCl a thick white solid precipitated, pH was slightly acidic (~3-4 vs indicator paper). The mixture was briefly heated to boiling in the microwave oven[3] and filtrated through 3 coffee-filters. The solids were washed twice with 100 ml near-boiling distilled water, allowed to cool and stand overnight, absorbed water was removed by gentle squeezing the filter, and the solids were dried in an oven under an air flow at gentle heat (70°C).
Yield : 22.27 g crude terephthalic acid as off-white solid

It was noted that the hydrolysis proceeded more swiftly at the moment NaOH was added, confirming that a mixture of KOH and NaOH increased the rate of hydrolysis compared to KOH or NaOH alone.

terephthalic_acid_crude_wet.png - 536kBterephthalic_acid_crude.png - 1.1MB


Experiment 2
In a 500 ml screw-top glass jar were placed 23.40 g PET[1], 8.60 g KOH, 8.20 g NaOH and 17 ml demineralised water. The contents were heated on a boiling water bath for 1 hour, after which another 8.00 g NaOH and 17 ml water were added, and heating continued for several hours until nearly all PET went into solution. Subsequently, 250 ml boiling water was added, mixed well, and filtrated through 2 coffee-filters. The dried undissolved fraction weighed 2.23 g.

After gentle acidification portionwise with azeotropic aqueous HCl, white solids precipitated, which were filtrated off and washed 2x with 100 ml boiling water, and dried in the oven in a stream of air at gentle heat (70°C).
Yield : 12.70 g crude terephthalic acid as off-white solid


Recrystallisation according to US patent 3465035
In a 1 liter beaker, 41.02 g (0.5 mol) anhydrous sodium acetate was dissolved in 100 ml demineralised water with gentle heating and stirring. Next, 9.53 g 80% acetic acid was added, followed by demineralised water up to the 500 ml mark. This solution was brought to a boil and, with good magnetic stirring, 15.49 g crude terephthalic acid was added in small portions. The resulting solution was allowed to cool slowly with gentle stirring on the hotplate to room temperature. On cooling needle-shaped crystals precipitated. These were filtrated, the filtrate saved[4], the solids washed 2x with 100 ml boiling water and left to air-dry.
yield : 7.05 g terephthalic acid as free flowing crystals

The patent did at one point mention a preference for pre-saturating the buffered recrystallisation solvent with terephthalic acid. A second experiment, using the filtrate of the first recrystallisation, yielded a better result: 10.45 g free flowing terephthalic acid from 15.50 g crude.

Acidification of the filtrate after this yielded another 3.42 g terephthalic acid, but of lower quality.

terephthalic_acid_recrystallised.png - 414kB


Notes
[1] clear packaging material, cut into ~ 0.5 cm x 0.5 cm pieces
[2] weighed on a kitchen scale with +/- 1 g accuracy
[3] 10 seconds at 1000W
[4] the filtrate can be reused on another batch as long as pH is maintained between pH 4.0-5.5

palico - 25-3-2024 at 07:27

Dear dicyanin, thanks for your experiment, but how have you checked the identity and purity of your products ?

Texium - 25-3-2024 at 12:38

Quote: Originally posted by dicyanin  
Experiment 1
In a 500 ml screw-top glass jar were placed 25 g PET...
Yield : 22.27 g crude terephthalic acid as off-white solid
Even if you started with completely pure PET, that would be a 103% yield. I think "crude" is likely understated, and you may have a significant amount of oligomers present in that sample.

kmno4 - 26-3-2024 at 09:31

Quote:
Even if you started with completely pure PET, that would be a 103% yield. I think "crude" is likely understated, and you may have a significant amount of oligomers present in that sample.

Has anyone told you lately, that you are boring ? Instead of writing 0-value posts, maybe you should simply take a walk ?

dicyanin - 27-3-2024 at 10:56

Quote: Originally posted by palico  
Dear dicyanin, thanks for your experiment, but how have you checked the identity and purity of your products ?

Both the crude and recrystallised terephthalic acid, despite difference in appearance, showed a clean single spot on TLC.
Attached are 2 chromatograms, crude and recrystallised TPA compared to oxalic acid and benzoic acid. The latter two showed streaking as the concentration of the sample must have been too high, but still their Rf values are widely different compared to TPA. Apparently terephthalic acid's "slightly soluble in ethanol" is just the right concentration for a perfect TLC spot.

PET_hydrolysis_TLC_oxalic.png - 395kB PET_hydrolysis_TLC_benzoic.png - 484kB


Quote: Originally posted by Texium  
Quote: Originally posted by dicyanin  
Experiment 1
In a 500 ml screw-top glass jar were placed 25 g PET...
Yield : 22.27 g crude terephthalic acid as off-white solid
Even if you started with completely pure PET, that would be a 103% yield. I think "crude" is likely understated, and you may have a significant amount of oligomers present in that sample.

Crude terephthalic acid seems to hold up a lot of solvent compared to the free-flowing recrystallised form, I suspect even after prolonged drying in the oven it wasn't fully dry yet.
Another possibile explanation is that in that experiment the 25 g PET was weighed on a kitchen scale with +/- 1 gram accuracy, while all other measurements were done using a more precise scale with +/- 0.01 gram accuracy.

According to the attached Yoshioka et al. 2003 paper, where they studied the hydrolysis of bottle waste grade PET by prolonged heating with concentrated aqueous NaOH, the main impurity encountered was oxalic acid, as ethylene glycol in concentrated alkali at elevated temperatures reacts with oxygen in air, but under those conditions sodium terephthalate was stable to oxidation, and the terephthalic acid yield was 100 mol% under all conditions.

Hence why I did a second TLC to check for possible oxalic acid presence in the crude sample.

The Yoshioka paper notes that oligomers were only encountered in direct (acid-catalysed) transesterification process using glycols.

Also attached is Abdelaal et al. 2008 paper where they study the degradation of PET at elevated temperature but using dilute (5%) aq. NaOH. Main points of interest from that paper, is that addition of catalytic Zn2+ or Ca2+ significantly increases the reaction rate of the hydrolysis, but their most interesting finding is that addition of a small amount of phase-transfer catalyst (tetraethylammonium chloride) tripled the reaction rate.

Attachment: used_PET_alkaline_hydrolysis_NaOH__yoshioka2003.pdf (88kB)
This file has been downloaded 45 times

Attachment: chemical_degradation_of_PET___abdelaal2008.pdf (230kB)
This file has been downloaded 44 times

[Edited on 27-3-2024 by dicyanin]

palico - 27-3-2024 at 14:50

dear dicyanin, i see you are a great chemical experimentalist. I do not so much on your TLC plate, the image quality is not good, but I believe in your results. I wonder also, PET oligomers would probably not move at all and remain at Rf = 0, in TLC and eluent system.
Try a proton NMR in DMSO. In my location, such analysis cost no more than 20 euro.

dicyanin - 28-3-2024 at 00:45

Thanks for your kind words palico. You are quite right about the quality of the pictures, my only excuse is that it wasn't easy fiddling about in total darkness with the only light source coming from 2x 5W UV-C tubes, trying to take pictures with a potato tier phone whilst wearing dark-tinted eye protection :D.

However, practice makes perfect, this morning I tried again and took these pics. To my shame I must admit yesterday I mistook a wire shadow for a benzoic acid streak yesterday, the actual spot is near the solvent front (shown by the pencil tip).
oxalic_TLC.jpg - 156kB benzoic_TLC.jpg - 110kB

I also calculated the following Rf values :

terephtalic acid : 0.75
benzoic acid : 0.83
oxalic acid : 0.10

Attached paper (Kolesinska et al. 1966) using the same solvent system (EtOH 96%/NH3 25%/H2O 5:2:2) found following Rf values :

benzoic acid : 0.80
oxalic acid : 0.36

Only the oxalic acid was off compared to my results. However, using different proportions (EtOH 96%/NH3 25%/H2O 5.3:4:0.7) they found :

benzoic acid : 0.75
oxalic acid : 0.05

so it seems that altering the ammonia:water ratio affects the oxalic acid Rf but not so much the benzoic acid Rf. As my source of ammonia is of a technical grade rather than an analytical one, I suspect that may be the reason for the deviation.

Attachment: benzenecarboxylic_acids_TLC__Kolesinska1966.pdf (664kB)
This file has been downloaded 40 times

palico - 3-4-2024 at 21:30

Dear detto:

indeed, by pyrolysis of PET it is possible to get benzene directly. I performed this experiment and opened a discussion, you can read here.

But, please, do not go off topic anymore.

Hexabromobenzene - 8-4-2024 at 13:44

Terephalic acid is of little interest. In theory, p-hydroxybenzoic acid or p -xylenediol by electrochemical reduction can be made from it. The p-diaminobenzole can also be made from this acid

However, these compounds are of little interest. Are there any partial decarboxylation methods of terephtallic acid into benzoic acid? For example, pyrolysis sodium terephtalate sodium with lime. Or isomerization to o-phthalic acid

bnull - 8-4-2024 at 19:20

Quote: Originally posted by Hexabromobenzene  
Are there any partial decarboxylation methods of terephtallic acid into benzoic acid? For example, pyrolysis sodium terephtalate sodium with lime. Or isomerization to o-phthalic acid

I wonder if even more interesting products could be made by a Borodin reaction using silver terephthalate and a halogen (chlorine or iodine). An intramolecular ester, for example, or dihalobenzene, or even the wossname (tricyclo[4.2.2.22,5]dodeca-1(8),2,4,6,9,11-hexaene) in the figure below. Sigma-Aldrich sells 100 mg of the carboxylic acid derivative of the wossname at $110 (https://www.sigmaaldrich.com/US/en/product/aldrich/s52559#pr...).
cyclophane.png - 6kB
Edit: As @Texium has pointed out below, what Sigma-Aldrich sells is a derivative of [2.2]paracyclophane, and not the long-named compound I have drawn previously. The structure above is the correct one.

There's also 4,4-carbonyl dibenzoic acid.
4 4-Carbonyldibenzoic Acid.png - 10kB

Alkaline hydrolysis of clear PET from a water bottle

And the latest developments of my ongoing experiment with alkaline hydrolysis (using the recipe presented in [1]) are quite interesting too. The reaction has sped up about 3 days ago. The pieces of PET that were slowly dissolving some weeks ago are for the most part gone, and the remaining few look like thin shredded plastic film. There's a deposit of disodium terephthalate at the bottom of the polypropilene vial.

I took some pictures and, despite the quality being too low, I'll share them as promised about a month ago. They were taken today. The coin is a 10 yen coin (23.5 mm in diameter). The engravings in its reverse help to focus the bloody camera Samsung has been using in some of their smartphones.

The first and second pictures, taken about one month apart, show the deposit of salt (presumably disodium terephthalate) at the bottom of the saturated solution. Notice the existence of two layers with different colors. The upper layer's yellowish tint is due to a speck of crap that fell into the vial when I was filling it and dissolved right away before I could curse or remove it. The lower layer, as white as boric acid, appeared after the rate of hydrolysis increased. The solution smells funny, quite like glycerol.
20240306_101241.jpg - 45kB 20240409_205351.jpg - 65kB

The third picture shows a crystaline mass that had formed in a corner of one of the PET pieces. I reported its formation about a month ago. It has grown to the size shown since then.
20240409_093526.jpg - 124kB

The fourth picture shows the crystaline mass after maceration in ethanol azeotrope. The alcohol was heated almost to boiling and tiny crystals, not discernible in the picture thanks (again) to the guys at Samsung, have precipitated on cooling. The inset on the upper right corner can give a rough idea.
20240409_205635.jpg - 67kB

The fifth picture shows a badly attacked piece of PET floating in water. The surface is finely grooved and some regions were torn apart or shredded by the solution. I removed it as a sample after seeing it was the first one to "give up", about one month ago. The fourth picture shows the same piece of polymer in a darker background.
20240409_211442.jpg - 84kB 20240409_211703.jpg - 100kB

Laziness and a splitting headache prevent me (for now) from going further and precipitating terephthalic acid from an aqueous solution of the salt. A previous test with a small sample (10 mg or so of salt) showed the formation of a milky emulsion, with particles so small that the time they took too to settle down was beyond my boredom limits. I'll try to rig a centrifuge tomorrow or the next day and do it.

I don't have an NMR spectroscope, or access to one, for that matter. My laboratory has been in a regretful state of disrepair for almost a decade, concomitantly with my being broke most of the time.

[1] Sibel Ügdüler et al. Towards closed-loop recycling of multilayer and coloured PET plastic waste by alkaline hydrolysis, Green Chem., 2020, 22, 5376-5394. https://pubs.rsc.org/en/content/articlehtml/2020/gc/d0gc0089...

[Edited on 10-4-2024 by bnull]

Texium - 9-4-2024 at 05:54

Quote: Originally posted by bnull  
or even the wossname (tricyclo[4.2.2.22,5]dodeca-1(8),2,4,6,9,11-hexaene) in the figure below. Sigma-Aldrich sells 100 mg of the carboxylic acid derivative of the wossname at $110 (https://www.sigmaaldrich.com/US/en/product/aldrich/s52559#pr...).
The structure you linked to on Sigma is called [2.2]paracyclophane (the structure you drew is missing the ethylene bridges between the rings). One of my coworkers has actually been trying to access these starting from terephthalaldehydes for the last 3 years and is close, but still not quite there. Seeing the grief he’s gone through even with excellent resources at his disposal, I’d say it’s not an amateur-friendly synthesis.

palico - 9-4-2024 at 20:22

Quote: Originally posted by Hexabromobenzene  
Terephalic acid is of little interest. In theory, p-hydroxybenzoic acid or p -xylenediol by electrochemical reduction can be made from it. The p-diaminobenzole can also be made from this acid

However, these compounds are of little interest. Are there any partial decarboxylation methods of terephtallic acid into benzoic acid? For example, pyrolysis sodium terephtalate sodium with lime. Or isomerization to o-phthalic acid


The interest of terephtalic acid is to make PET. The PET hydrolysis I have presented is a form of chemical circular recycling, which transforma a waste in its precurso material, from which new virgin PET can be made. Mechanical recycling is instead difficult and resulting PET has different physico-chemical properties than starting one.

@bnull: thanks for your contribution. Yes that is paracyclophane, it is quite hard to synthesize.