Sciencemadness Discussion Board

Chemistry of Terephthalic acid

 Pages:  1  

Boffis - 2-8-2015 at 14:10

The recovery of terephthalic acid from PET bottles is now well described elsewhere on this forum and the topic of what can be done with the resulting acid has been broached occasionally in various threads. Unfortunately, these threads usually take the form “can I make xxxx from PET bottles?” rather than looking at what can reasonably be prepared from the free acid.

Has anyone out there ever prepared any derivatives from terephthalic acid and is prepared to share their results with us?

I recently ran a series of experiments on the alkali hydrolysis of PET resin from water bottles and now have roughly 200g of terephthalic acid and sodium terephthalate. By carrying out the hydrolysis in batches of 50-60g of PET at a time and recycling the liquor into the next batch I managed to achieve recoveries of >95% of a high quality product. The preparation of solid, crystalline di-sodium terephthalate proved tricky since it appears that the solubility/temperature gradient is almost flat and so a saturated solution of the sodium salt does not crystallise at all. The best method was found to be to evaporate the sodium salt solution on a water bath until a slurry of sodium terephthalate crystals had formed and filter.

What next?
Having prepared the acid and sodium salt I decided to investigate its chemistry. I have not yet had chance to investigate it’s chemistry with practical experiments but in general it appears that terephthalic acid is very stable and not very reactive and direct substitution is difficult and limited. After a moderately extensive search through the literature I came to the conclusion that there are 6 routes to derivatives of terephthalic acid that are reasonably achievable to amateurs though not necessarily amateur friendly or desirable. So here are some ideas for anybody that has contemplated trying to prepare something from terephthalic acid but was afraid to try:

1) Salts: the di-sodium terephthalate is easily prepared and from this can be prepared a wide range of generally sparingly soluble salts by double decomposition. Some the rare earth salts are reported to have fluorescent and luminescent properties (1), (2).

2) Esterification: Terephthalic acid can be converted to its dimethyl or diethyl ester with the appropriate alcohol and either sulphuric acid or gaseous HCl (5, 7). The resulting esters are crystalline solids. These esters are fairly easily half hydrolysed (5, 7, 9) to terephthalic acid monomethyl/ethyl ester and it may be possible to generate the half amide by a modified process too, from which, via Hofmann’s degradation, 4-aminobenzoic acid ester may be available. The mono esters can also be converted to monoacid chlorides (7) and from these via treatment with ammonia the monoamide. There is an SM thread by Picric-A on the this process but unfortunately it is lacking details of experimental procedures and he failed to provide any in spite of repeated requests from other members to do so and his procedure from ammonium hydrogen terephthalate must remain questionable, however, see the comments below. The 4-aminobenzoate ester is easily nitrated and the nitration products reduced with sulphide to 3,4-diamine and 3,4-diamino-5-nitrobenzoate esters which when treated with nitrous acid are converted to benzotriazole derivatives (my interest). Other possible routes from the 3-nitro-4-amino ester include hypochlorite oxidation to a benzofuran derivative.

3) Nitration: Nitration is reported to require rather forcing condition with large excesses of fuming sulphuric and fuming nitric acid to yield the mononitroterephthalic acid (only one isomer is possible) (3, 8). However, I think it would be worth trying this nitration with alkali nitrates and conc. Sulphuric acid at moderately elevated temperatures instead. The reduction of the nitro compound to the amino compound and then via diazotization to the hydroxyterephthalic acid has been reported (3); Wegschieder (8) also described the reduction of nitroterephthalic to the amino acid (effectively an anthranilic acid derivative). The presence of the amino or hydroxy group may be sufficient to activate the nucleous to further substitution and it is interesting to speculate on the possible de-carboxylation on further substitution in the manner of salicylic acid to picric acid under trinitration to yield in this case 2,4,6-trinitro-3-hydroxybenzoic acid. It is possible that both carboxylic acid groups would be lost in this process as the reference above (3) reports that bromination of 2-hydroxyterephthalic acid occurs readily forming 2,4,6-tribromophenol.

4) Mercuration: Terephthalic acid can be mercurated with mercuric oxide or mercuric acetate and then converted to the monohaloterephthalic acid by treatment with the appropriate halogen (4). While the process sounds simple enough the very long refluxing (>10 days!!) and the production of much unusable mercury bearing organic waste makes this process impractical as a preparative route.

5) Iso and terephthalic acids can be converted to their respective acid chlorides and then reduced to the corresponding dialdehyde (6). This route is also of limited value to amateurs in view of the need for compounds such as thionyl chloride or phosphorus pentachloride.

6) Another possible routes to derivatives is the direct amidation by treatment of the free acid or ammonium salt with urea and sulphamic acid (10 but the reference is not entirely clear as to whether the product is the mono or diamide, the former is implied but no supporting data is provided) or with urea and boric acid (11) to the half or full amide. It may then be possible to convert the amides into the nitrile with ammonium sulphamate (12) to give the 4-cyanobenzoic acid or 1,4-dicyanobenzene respectively.

The decarboxylation of terephthalic acid to benzene or benzoic acid has been discussed on this site in several threads but I have found no published article specific to terephthalic acid for this process yet.

(1) Dalguebonne et al., Inorg. Chem., (2008), v47(is9) p3700. DOI; ic702325m (luminescent RE salts)
(2) Dalguebonne et al., Inorg. Chem. (2006), v45(is14) p5399. DOI; ic060241t (Erbium salts)
(3) Burkhardt GA. Berichte., (1877), v10, p145. (hydroxyterephthalic acid via nitration, reduction to amine and diazotization)
(4) Frank et al., JACS (1929), 51(9), pp2785- . DOI; ja01384a027 (mercuration)
(5) JOC (1960), v25(5), p817; DOI jo01075a039 (esterification and half hydrolysis)
(6) Rosemund and Zetsche, Berichte., (1921), v54, p2888. (reduction of iso and terephthalic acid to appropriate aldehyde via the acid chlorides)
(7) Cohen & Pennington, JCS, (1918), v113, p63. (diethyl ester, diacid chloride, monoethyl ester and monoester chloride of terephthalic acid and nitrophthalic acid).
(8) Wegschieder R., Monatsch., (1900), v21, p621. (nitration of terephthalic acid and its salts)
(9) Wegschieder R., Monatsch., (1902), v23, p405. (nitroterephthalic acid esters and half esters)
(10) US 3663589 J Luecke, Process for the production of nitriles (-from carboxylic acids with sulphamic acid see example 4)
(11) SM thread “Benzoic acid, Carbamide>benzamide>aniline” http://www.sciencemadness.org/talk/viewthread.php?tid=4201
(12) Can J Research, J. L. Boivin, (1950), v28, p671, (Nitriles from amides & sulphamic acid)


Pumukli - 3-8-2015 at 03:35

Ta-da!

Congrats Boffis! :-)

Your quite detailed work is warmly appreciated, thanks!

PET hydrolysis and terephtalic acid chemistry is something that bothers me a lot. It is so OTC.
Now you compiled a very good starting point to help me (and probably others) to pursue this route.

How do you cut up the bottles? With a sharp knife? Or do you have any special technique? How small the bits are? How long doeas it take to hydrolise them?

Nicodem - 3-8-2015 at 04:07

Boffis, thanks for the excellent review. This forum needs more such topics.

I can add that terephthalic acid is the prototypical ligand for the synthesis of metal-organic framework materials (MOF), a class of coordination polymers.
Two randomly chosen (application oriented) review articles: DOI: 10.1021/cr9003924 and DOI: 10.1021/cr200190s

I'm often thinking why home chemists show practically no interest in coordination polymers. They appear so suitable for a few reasons. Some are very easily synthesized (at least microcrystalline and solvated), there is a lot yet to be discovered (at least application-wise), their structure is often very hard or impossible to determine, so that characterization by application is acceptable even for publishing... And if you go on to synthesize your own multidentate ligands, then you are bound to learn a great deal of organic synthesis as a bonus.

papaya - 3-8-2015 at 05:08

Can 4-aminobenzoic acid (which probably could be made from terephtalic acid) self-condense to long chain linear polyamide ? If yes, supposedly the polymer is conductive.

Boffis - 3-8-2015 at 12:20

@Pumukli, I found that the larger 1.5-2 litre soda bottles are best but I currently use the 600ml water bottles because we are handed a couple of these every day on site. I remove the labels, which are just a plastic band, and then remove the neck and the thick crown at the base with a knife because they are difficult to cut up and then chop up the remainder with a pair of scissors. I cut it into strips about 1cm wide and then into small squares of about 1cm2, it takes ages to cut up 250g! I wondered if you could use a domestic blender once you had removed the top and bottom but I didn't try it.

@Nicodem, its funny you mentioned the coordination polymers. I had never heard of them before I read the two papers on the rare earth compounds but they are almost exactly what I was originally chasing. My original target was 2,5-dihydroxyterephthalic acid, since salicylic acid produces a nice purple but extremely soluble complex with ferric iron I wondered if two salicylic acid groups on the same benzene ring might produce a similar but less soluble complex by virtue of its desire to form 2 or 3D structures. However, my literature search seems to indicate that the dihydroxyterephthalic acid is generally prepared either from 2,5-dibromo-p-xylene or via a double Kolbe reaction from hydroquinone (salicylic acid is produced via the analagous reaction from phenol). I am actually quite surprised that an isolated carboxylic acid group can form sufficently stable coordination compounds to build such 3D structures, I would have thought that the extra OH adjacent to the carboxylic acid group would have greatly improved stability; maybe it does have I haven't tried it yet.

Incidentally; will researching terephthalic acid I stumbled across a paper that covers another topic that crops up regularly., that of preparing phthalic acid by oxidizing hydrocarbons. I will have to look for the paper again but in essence it described the preparation of all three phthalic acid isomer by oxidation of mixed isomer xylene and then separation of the three resulting acid by fraction solution and then crystallisation of the barium salt. They enriched there feed stock by partial removal of the o-isomer with a 1.5m tall fractionating column (!) first but I think the process is capable of manipulation to allow the use of "as bought" xylene to yield all three acids. I am looking into this at present.

@papaya, I'm sure its possible to polymerise it but it might be rather difficult. I don't know much about these things but maybe you could use the techniques they use to create peptides from amino acids.

Waffles SS - 4-8-2015 at 10:45

Thanks Boffis,

I tried Esterification of Terephthalic acid with Short-chain alcohols like Methanol , Ethanol and i have to say this is not possible by standard method.

Finally i made esters of Terephthalic acid by acyl chloride route.

This is interesting for me to know:

P-Phenylenediamine from Terephthalamide is possible?

Does Hofmann rearrangement is possible for both amide in Terephthalamide at same time?


[Edited on 4-8-2015 by Waffles SS]

Pumukli - 4-8-2015 at 11:08

A New Process for the Production of p-Phenylenediamine Alternatively from Polyester Waste, Terephthalic Ester, or Terephthalic Acid

Hans G. Zengel , Manfred J. Bergfeld
Ind. Eng. Chem. Prod. Res. Dev., 1976, 15 (3), pp 186–189
DOI: 10.1021/i360059a009
Publication Date: September 1976

If someone could access (and post) it that would be nice! :-)
I can't.

Boffis - 5-8-2015 at 08:12

@waffles; I am surprised you found it so difficult to esterify terephthalic acid by the Fischer process, it is described in several of the references above so I would give it another go. I will certainly try it myself shortly.

By the way what method did you use to generate the acyl chloride? It looks like Pumukli's references answers your question about hofmann degradation of terephthalamide.

@Pumukli, attached is you reference, very interesting indeed. It sounds as though only a small over pressure of ammonia is required so it may be possible to rig up such equipment at home.

Attachment: Production of p-phenylenediamine from terephthalate esters Ind Eng Chem Prod Res Dev Zengel & Bergfeld 1976.pdf (449kB)
This file has been downloaded 885 times

[Edited on 5-8-2015 by Boffis]

Waffles SS - 5-8-2015 at 09:06

@Boffis,

I made Terephthaloyl chloride by refluxing excess Thionyl chloride with terephthalic acid and then distilling the product.

1.JPG - 77kB
2.JPG - 65kB
3.JPG - 69kB
4.JPG - 78kB
5.JPG - 64kB


I also sent requested @Pumukli reference by private message to him according to copy right law ;)

[Edited on 5-8-2015 by Waffles SS]

Pumukli - 5-8-2015 at 09:49

I'm glad that I may helped you with the reference. Thanks, I got it, read it.

Regarding the first stage, the ammonolysis, it works as I expected (silently). My original idea was that PET is an ester and esters usually give amids when reacted with NH3. I hoped that terephtalic acid is no exception and glycolic esters are also working in this regard.

Now the overpressure issue.
I think it may be circumvented by using much finer starting material. Instead of PET chips I plan making PET "grindings" with some sort of low speed grinder machine. It may also take ages to produce a sizable ammount this way but the much increased surface area may outweigh the sweat and tears. Ball milling with dry ice would probably be even better, but is not realistic for me.


papaya - 5-8-2015 at 15:45

Quote: Originally posted by Pumukli  


Regarding the first stage, the ammonolysis, it works as I expected (silently). My original idea was that PET is an ester and esters usually give amids when reacted with NH3. I hoped that terephtalic acid is no exception and glycolic esters are also working in this regard.



Once when storing 15% ammonia in a PET bottle it "corroded" away just in a week, nearly up to making holes in it and all covered in white specs ,especially ABOVE the level of liquid ! As ammonia is not as "basic" as say NaOH, can I speculate that what formed was amide(s)?

Waffles SS - 5-8-2015 at 19:59

Quote: Originally posted by papaya  


Once when storing 15% ammonia in a PET bottle it "corroded" away just in a week, nearly up to making holes in it and all covered in white specs ,especially ABOVE the level of liquid ! As ammonia is not as "basic" as say NaOH, can I speculate that what formed was amide(s)?


Yes,

Ammonolysis of esters lead to Amides

Aminolysis of PET by Methylamine and Ethylenediamine at room temperature going faster than Ammonia and even faster if you use small amount of Phase Transfer Catalyst like TBAB or CTAB(google it to find interesting articles)



[Edited on 6-8-2015 by Waffles SS]

Boffis - 6-8-2015 at 10:34

@Waffles, Nice to have access to thionyl chloride. I might try with PCl5 to give terephthalylchloride and phosphorus oxychloride. But with the number of published descriptions of esterification of terephthalic acid with methanol or ethanol and sulphuric acid it must be possible. I am going to try it with anhydrous methanol and possibly azeotropic water removal after a reasonable reflux period.

Reading the Zengel and Bregfeld paper again, they refer to the glycol not only as a solvent also as a catalyst. This and Papaya's comments make me wonder if you could just "oil" the PET chips with glycol and pack them in a glass vessel and pass ammonia gas through it slowly over a period of a week or two at atmospheric pressure. Or possibly using a Schott type bottle in an oil bath at 140 C for just a day or so.

Waffles SS - 6-8-2015 at 10:56

Quote: Originally posted by Boffis  
@Waffles, Nice to have access to thionyl chloride. I might try with PCl5 to give terephthalylchloride and phosphorus oxychloride. But with the number of published descriptions of esterification of terephthalic acid with methanol or ethanol and sulphuric acid it must be possible. I am going to try it with anhydrous methanol and possibly azeotropic water removal after a reasonable reflux period.


You can use S2Cl2 instead of Thionyl chloride(also i have access to Riedel-de Haën S2Cl2 :D)

Esterification of Terephthalic acid with methanol or ethanol is possible if you can bring out formed water from reaction
This is difficult because Methanol and Ethanol have lower bp than water(my suggestion is using Molecular sieve 3A or using large amount of Sulfuric Acid as water absorber)


[Edited on 6-8-2015 by Waffles SS]

Boffis - 6-8-2015 at 13:16

@Waffles, one way to get the water out is to use a large excess of dry ethanol. After refluxing for some time the excess alcohol is distilled off, it will remove the water as the azeotrope ethanol-water mixture. I don't think this works with methanol.

It may also be possible to remove the water by co-distillation with an immisible solvent such as hexane.

The paper cited above suggest that after an adiquate reflux period conversion reaches at least 70% ester and since the unreacted acid is recoverable I don't see am big problem.

Waffles SS - 6-8-2015 at 21:12

Use Chloroform for making ternary azeotrope(Water + Methanol + Chloroform=b.p.azeo. 52˚C)
https://en.wikipedia.org/wiki/Azeotrope_tables

I usually use below apparatus for making this type of ester and another Methyl or Ethyl ester(Oxalate,Tartarate ,..)


untitled.jpg - 36kB

At end you should wash your ester with Sodium Carbonate solution for removing unreacted acid and mono ester of Terephthalic acid


[Edited on 7-8-2015 by Waffles SS]

Boffis - 8-8-2015 at 07:11

@Waffles; interesting idea. I was planning to try n-hexane as a substitute for benzene. My only worry about chloroform is how reactive is it towards the sulphuric acid-alcohol mixture?

I have the equipment to use the zeolite assisted esterification but how much zeolite do you need per gram of water that will be formed?

Waffles SS - 10-8-2015 at 05:29

Use PTSA(Para-toluenesulfonic acid) instead of sulfuric acid as catalyst

I dont have experience on zeolite but i prefer Molecular sieve

Boffis - 19-8-2015 at 13:28

@Waffles; most molecular sieves ARE zeolites!

I thought I would have a go at nitrating terephthalic acid today, here is what I did;-

I added 4g of dry terephthalic acid to 20ml of 96% sulphuric acid and then added 2.5g of finely ground dry sodium nitrate. After mixing I let it stand for 30 minutes but nothing seemed to happen so I put it on a hotplate that I had used earlier but which was still warm. After another 30 minutes to temperature was such that the beaker was uncomfortably hot to the touch but still nothing appeared to happen so I turn the hotplate on again and heated it further until the the mixture began to to simmer and give off small amounts of nitric acid smelling vapors but still not visible reaction! At this point I gave up and left the mixture to cool down. When almost cold I scraped the viscous white mass into 100ml of cold water. The result was an immediate bright yellow colour and much insoluble material, the whole resembled a suspension of sulphur powder!

I heated the mixture almost to boiling to try and dissolve the precipitate but to no avail, so I cooled it again and then filtered it. The filtrate is a beautiful bright yellow colour but the solid is white and looks like the original terephthalic acid. I will investigate the solid further but I could not recover anything from the yellow acid residue.

Waffles SS - 19-8-2015 at 20:33

Sorry you are right

I didnt know my 3A molecular sieve made of zeolite

[Edited on 20-8-2015 by Waffles SS]

Boffis - 3-9-2015 at 14:08

I was so intriged by Papaya's comments about the "corrosion" of a PET bottle by aqueous ammonia I just has to try it. The results were slow but amazing; no pressure vessel, no gaseous ammonia or high temperatures just patience!

I took the thicker parts of a PET water bottle from the top and bottom to see how far the process would penetrate. I smashed them with a hammer to make them fit through the top of a glass soda drink bottle but would wedge in the top out of contact with the liquid, I would estimete the weight to be 3 to 4g. I then added 50ml of 32% ammonia solution so that the plastic was wetted but most of the liquid drained to the bottom of the bottle and replaced the lid. I then left the bottle on a shelf in my office for 3 weeks while I was on holiday. When I returned the plastic had turned white and partly crumbled and all was lying in the ammonia liquid at the bottom of the bottle. On shaking much of the white plastic was reduced to white crumbs which suggests that in the thinner sections were completely altered though the thick disc at the base and the lip on the neck are not completely altered.

PET 12 amide formation.jpg - 454kB White altered PET after 3 weeks

I will let it stand for a further 3 or 4 weks and then filter off the solids, wash them and then gently grind them to see how much PET is left.

I think that if 15% ammonia solution works then you probably only need a 30-40% excess af ammonia solution at 32%. This coupled with the use of only the thinner wall sections of the PET bottles could be an interesting "home" route to various compounds. You just need to be patient.

Morgan - 3-9-2015 at 14:45

I recall hearing something about PET bottles that save costs now by using double layers, one of a less expensive material.
Here's some article about special purpose PET bottles that use multi-layers, just on the off chance the ammonia treatment produces different results at times with various PET samples.
http://www.ptonline.com/articles/barrier-pet-bottles

Another
http://www.gasbarriertechnologies.com/petbottles.html

[Edited on 4-9-2015 by Morgan]

papaya - 4-9-2015 at 06:10

Good to hear about your success Boffis, what I also remember from my accident is that I didn't want to throw away that ammonia which stayed in PET bottle considering it must be nearly pure, so I transferred solution into the glass bottle. After a day a thin layer of well formed needle-like crystalls precipitated from totally clear ammonia solution! Seems that, whatever product it gives, it is only marginally soluble in water, and it may be that what you obtained after weeks of reaction in form of white crumbled stuff is not the original plastic anymore, but the product!
One more idea at the end: pet bottles melt when you heat them. When solidified it's no more elastic, but can be crumbled/milled easily into powder, which must reduce reaction time down to several hours/days.

[Edited on 4-9-2015 by papaya]

PHILOU Zrealone - 6-9-2015 at 07:07

Thank you Boffis for this excellent work!
Here are my 50 cents... since I also have thought for years about the possible uses of PET and terephtalic acid. ;)

1°) If one has access to NaN3 or to N2H4 he could perform:
°Cutrius reaction via Ar-CO-NH-NH2 and HNO2
°Schmidt reaction with NaN3 and H2SO4
Partial reactions would lead to HO2C-Ar-NH2 and complete reaction would lead to H2N-Ar-NH2
From there diazotation is possible with a lot of possible products.

2°) H2N-Ar-NH2 could be reacted with Cl-CO-Ar-CO-Cl (terephtalyl chloride) or with MeO-CO-Ar-CO-OMe (dimethyl terephtalate) to yield kevlar (-NH-Ar-NH-CO-Ar-CO-)n

3°) H2N-Ar-NH2 reacted with aqua regia (HNO3/HCl) will yield chloranile (O=C(-CCl=CCl-)2C=O), with HBr/HNO3 bromanile and probably with I2/HNO3 iodanile.
The halogens are very labile and may lead to halogen exchange reaction; for example tetrabromo-p-quinone (bromanile) will exchange 2 I (diiodo-dibromo-p-quinone) upon reaction of KI/EtOH and finally 4 I (iodanile) if longer exposure time, heat, reflux and more reactants.
The halogen exchange of chloranile, bromanile and iodanile is characterized by the easy replacement of each halide by azide group upon contact with NaN3 water solution forming the extremely dangerous tetraazido-p-quinone (intermediary stages are possible with mono, di, tri azides). This was covered by Axt in this forum.
Same is seen with NaNO2 what forms nitranilic acid (2,5-dihydroxy-3,6-dinitro-p-quinone) forming primary explosive salts like silver and lead nitranilates (the substitution passes maybe via a transcient tetranitro-p-quinone that hydrolyses spontaneously to the dihydroxy-dinitro compound).
Maybe other easy substitutions are possible like OH(-), C#N(-), NH3, NH2OH and N2H4 (this last one will probably reduce the quinone)...to test also amines like CH3-NH2, (NH2)2C=NH, ...

4°) Reaction of aqua regia or of HBr/HNO3 on terephtalic acid is not clear...
-maybe it remains unaffected
-or it is partially halogenated (and maybe decarboxylated) --> halo (dihalo, trihalo or tetrahalo) terephtalic acid and maybe polyhalobenzenes and polyhalobenzoic acid
-or decomposed into chloropicrin (Cl3C-NO2) or bromonitromethanes (CH2Br-NO2, CHBr2-NO2, CBr3-NO2).
The halopicrines are good ways to guanidine via reaction with concentrated ammonia solutions
Cl3C-NO2 -NH3 conc-> (H2N)2C=NH.HCl + NH4Cl + N2 + H2O
Chloropicrine may be dehalogenated and reduced by H2S in basic media to dichloronitromethane, chloronitromethane, nitromethane, methylamine and finally formaldehyde-methanol.

5°) Reaction of iodination (periodination) reagents (HNO3/I2, Cl-I, I2/SO3) allows one to get acces to iodo-, diiodo-, triiodo- and tetraiodo-terephtalic acid (this last one is a X-Ray contrast media for blood system or urinary system medical study).
Under harsh conditions decarboxylation may occure to yield pentaiodobenzoic acid and hexaiodobenzene...

6°) Hunsdieker reaction is not working well with terephtalic acid (less than 2% yield) but well with benzoic acid what will yield chlorobenzene or bromobenzene from silver benzoate (see point 7°) and Cl2 or Br2 in refluxing CCl4 or other suitable solvents.

7°) Thermolysis (425°C) of dry silver terephtalates generates a lot of products like benzoic acid, benzene, biphenyl as major products and terphenyl, quaterphenyl, quinquephenyl, 3,4-benzocoumarin as minor products.
See in requested documents "Thermal and Photochemical Decomposition of Silver Carboxylates", requested by me.

So from this tread, it is obvious many things (if not everything) can be done from easily got OTC PET...
The only limits are:
-your mind :)
-time :(
and as usual :D
-money :cool:


Boffis - 9-9-2015 at 08:12

@Papaya, yes I'm pretty sure you're right about the white product being the diamide and it also appears that your observation that moist ammonia gas is more aggressive than the solution is correct. With this in mind I am going to try some further larger scale experiments where the PET is isolated from the ammonia solution.

@Philou, thank you for the ideas you posted. There are some interesting possibilities here. I wonder if hydrazine solution would be more aggressive than ammonia solution? I have some 60% hydrazine hydrate solution (about 40% hydrazine), the aim would be to produce the dihydrazide directly from PET. The lower vapor pressure of hydrazine over this solution means you can probably reflux it at atmospheric pressure with minimal loss of hydrazine.

The direct halogenation would also be interesting if it can be made to work but given the inertness of terephthalic acid to nitration I think very harsh conditions will be required though once introduced these halogen are fairly reactive and can be replaced by OH using KOH.

Update on the ammonia experiment: After a vigorous shaking most of the white alteration product became detached from the residual PET but the the thicker parts of the threads and neck still remain see photo.

PET 13 amide formation 4wks.jpg - 326kB
The PET in ammonia after about 4.5 weeks, fragments of threads and the star from the base are clearly visible.

nlegaux - 11-9-2015 at 10:39

A reaction I haven't seen listed yet is Letts nitrile synthesis. If you react the terepthalic acid with stoichiometric amounts of KSCN, you would be able to produce 1,4-dicyanobenzene.

nlegaux

FriedBrain - 11-9-2015 at 12:23

Do you have any source or literature for that? I don't think that this reaction will take place at normal temperatures and pressure, but who knows?

aga - 11-9-2015 at 13:38

Nicodem, woelen, blogfast25 to name but a few.

Boffis - 11-9-2015 at 14:17

I've never heard of Letts synthesis before but having checked it out it actually look plausible particularly the lead thiocyanate modification. I wonder if you could apply it to the half ester to obtain a 4-cyanobenzoate ester. This reaction would be a more direct route to 1,4-dicyanobenzene than through the diamide.

Pumukli - 12-9-2015 at 11:05

Whoa, sorry for the following (almost) entirely offtopic comment, but I couldn't resist:

After Boffis mentioned Letts nitrile synthesis I also got curious and checked it on wikipedia. https://en.wikipedia.org/wiki/Letts_nitrile_synthesis
Then checked reference 5 and to my amusement the first real sentence of that article (after the usual intro and table of contents listings) went like this:

"Hydrogen cyanide was first prepared in 1782 by Scheele, who was later killed while attempting to isolate the anhydrous material."

Mmm, it was new to me that Scheele (who was a great chemist of his time) had died in a laboratory accident, by HCN.

nlegaux - 12-9-2015 at 14:27

It is rather ironic that something so seemingly foolish could kill such a brilliant chemist.

Another interesting reaction of carboxylic acids is the Kochi reaction, which allows for the production of alkyl chlorides from carboxylic acids using lead tetraacetate and the corresponding lithium halide.

nlegaux

[Edited on 9-12-2015 by nlegaux]

Boffis - 12-9-2015 at 16:35

@nlegaux, thank you for another interesting idea and another reaction I have never heard of. However, having delved into this reaction it would appear that if it worked on terephthalic acid the product would be p-dichlorobenzene. Thus to produce 1g of p-dichlorobenzene (cost £0.01 per g) would require about 8g (£3.00 worth of lead tetracetate); this is therefore analogous to using a solid gold hammer to turn your Rolls Royce, Porsche, Ferrari or whatever you drive into scrap metal.

I hope you don't feel that I am pissing on your camp fire but.....

nlegaux - 13-9-2015 at 12:32

It is no problem Boffis, just an idea I had (excellent analogy, by the way). I'm currently looking into performing the Kochi reaction on some Salicylic acid, so when I thought of the carboxylic acid groups on the terepthalic acid that was one of the first things that came to mind.

nlegaux

Boffis - 20-9-2015 at 14:27

OK I've stop my experiment described above with PET plastic and aqueous ammonia.

I shuck the bottle vigorously to seperate as much as possible of the white product from from the residual plastic and then strained the mixture through a plastic tea strainer to remove the coarse fragments. The white suspension was gravity filtered and then washed with copious amounts of water until the smell of ammonia had gone and the wash water was no longer alkaline to red litmus. The filter paper was then air dried. When dry the white powder readily seperated from the filter as a fine, soft white powder. I don't have scales where I am so I can't give yields but after 5 weeks all of the thinner side walls of the water bottles had completely disintegrated only the bottoms and necks remained. If the product proves to be terephthalamide this process represents a very useful route to 1,4-benzene derivatives from PET resin.

PET 14 residual pet.jpg - 722kB The residual pieces of the bottles

PET 15 terephthalamide.jpg - 1007kB The product; terephthalamide?

While I have no means at hand to test the material the only other two likely products are diammonium terephthalate and ammonium terephthalate monoamide but both of these are likely to be pretty soluble in ammonia solution and certainly the wash water. I didn't think to save the ammonia soution and try drying it to see if there was any soluble material.

When I get chance I will try a more scientific version of this with weighed quantities and better control.

aga - 20-9-2015 at 14:31

Inspiring ! Definitely have to try this at home, in the Lab.

[Edited on 20-9-2015 by aga]

nlegaux - 24-9-2015 at 06:15

If it is terepthalamide, you can perform a Hofmann rearrangement to produce phenylenediamine.

nlegaux

halogen - 24-9-2015 at 11:27

Could contain oligomers capped with ammonia, with some ethylene glycol remaining I mean.

Boffis - 24-9-2015 at 18:50

@halogen, yes very possibly. Its difficult to recrystallise give the low solubility in the solvent I have to hand so my plan is to repeat this with a pressure vessel that I have just aquired and see if I can get a product with the same properties/composition but in hours rather than months. I also intend to try the boric acid/urea process on my purified terephthalic acid. Unfortunately at the present time I am in the Cloud Forests of the eastern Andes so these experiments will have to waite until I am home again.

Boffis - 16-10-2015 at 13:08

Since we use ammonia on site I decided to repeat the ammonia decomposition at atmospheric pressure. I used a glass food jar with some polythene mesh supported on a ring of uPVC water pipe to seperate the PET fragments from the liquid aqueous ammonia solution (25%). I used only the thinner wall sections of the bottles, about 15g. I filled the jar up to the levelof the polythene mesh with about 70ml of ammonia solution and placed the PET in largish piece in the space above. I shuck the jar one to wet the PET and then let it stand for 3 weeks undesturbed. After this time the PET had become completely opaque and when I shuck the jar again the fragments completely disintegrated to white milky slurry. I filtered off the white solid but this time I preserved the filtrate. I washed and dried the white solid and will investigate this further.

The filtrate I placed in a shallow tray and left it to partially evaporate to remove most of the ammonia. By the evening only about 30ml remained and the smell of ammonia was very slight. A tiny amount of white solid precipitated as minute ponpoms. I then acidified the clear liquor with 10% HCl to about pH 1.5 (by test strips) and removed the white precipitate by filtration again. I estimate that about 0.5g of this material remained on drying. This material could be either terephthalic acid or the mono-amide or more likely a mixture of the two. Does anyone have a simple test to distinquish the two? I have some good quality filter paper and some solvents (70% ethanol, butyl acetate, THF etc) available here so chromatography may be possible.

Mesa - 17-10-2015 at 03:33

Add sodium hydroxide and hold a strip of wetted pH paper above it. If it turns blue it's an amide.

Paddywhacker - 18-10-2015 at 00:00

Quote: Originally posted by Boffis  
... I washed and dried the white solid and will investigate this further.
...


Did you do a melting point? Solubility tests?

Pumukli - 18-10-2015 at 03:05

Melting point of terephthalic acid is around 425 Celsius, its monoamide melts above 340 Celsius, diamide I don't know.
All I wanted to say is you will need a beefy melting point apparatus for sure. :-)

Paddywhacker - 3-11-2015 at 19:48

Wiki gives the sealed tube MP as 300 C and sublimation with decomposition at 402 C.

It should be possible to characterise it as either a form of PET or as the acid, ammonium salt or amide on the basis of simple tests.

[Edited on 4-11-2015 by Paddywhacker]

Boffis - 31-3-2016 at 10:14

Large scale ammoniolysis of PET resin over aqueous ammonia

Following my earlier small scale experiments exposing PET plastic to the fumes of aqueous ammonia and the reasonable success it appeared to have I decided to carry out a much larger scale experiment.

I set up an old desiccator with a piece of plastic mesh to replace the zinc plate and then placed 206g of plastic water bottles cut into rings so that they did not pack too densely. 350ml of concentrated 0.90 density ammonia solution were then poured over the plastic and allowed to drain into the reagent pit at the base of the desiccator. The mesh kept the plastic out of the ammonia solution and so apart from the initial drenching the plastic was only exposed to moist ammonia vapor. The rim of the desiccator was greased with vaseline and the lid was placed firmly in place and the whole was left in a cool dark corner in my outbuilding for nearly 5 months from early November 2015 until mid March 2016. I estimate that the average temperature during this period was about 7-8 C.

PET ammoniolysis 1.jpg - 988kB Desiccator after opening and draining the excess ammonia

PET ammoniolysis 2.jpg - 1.1MB The now white corroded PET resin

When the experiment was opened the whole of the content had turned completely white and opaque and it had settled considerably. The aqueous ammonia that was left was carefully decanted and the PET residue tipped into a large bowl. The desiccator was washed out with about 400ml of water into the bowl and the now very soft and fragile PET gently crushed to give a thick porridge like slurry. This was filtered using a large Buchner funnel; the filtrate was preserved for recovery of any free acids while the filter cake was dried for several days.

PET ammoniolysis 3.jpg - 848kB The washed and filtered amide

The dried cake was ground to a fine powder and dispersed into 800ml of hot water, simmered for a short while (about 10 mins) and then cooled. The fine white solid was again collected in a large Buchner funnel washed with about 300ml of cold water and dried. The yield of dried solid was 155g. It is highly insoluble in cold as well as hot water and it is not soluble in dilute sodium hydroxide. It is also very resistant to hydrolysis even by hot 40% sodium hydroxide. The nature of this material was not investigated further at present but the yield is very close to the theoretical yield of terephthalamide.

PET ammoniolysis 4.jpg - 1.1MB The rewashed amide product

The combined filtrates were acidified to Congo red paper with 18% hydrochloric acid to precipitate any free acids. The white precipitate was recovered by filtration at the pump and dried, it weighed 18.87g. The nature of this acid is uncertain but the two most like possibilities are terephthalic acid and its half amide, 4-amidobenzoic acid; some short chain partial hydrolysis products are also possible if they contain a free acid group.

The residual ammonia solution was partially distilled into fresh water to recover most of the remaining ammonia and the residue in the distillation flask (now rather cloudy but from which nothing settled out) was acidified with hydrochloric acid and the white precipitate filtered off, washed with about 50ml of water and dried yielded a further 1.002g of acid that was added to that already recovered, giving a toatl yield of 19.9g.

To determine the likely indentity of this compound it was decided to titrate the acid against standard sodium hydroxide using phenolphthalein as indicator. The choice of the indicator is important as the final solution should be sufficiently alkaline to neutralise the second base of any terephthalic acid. An immediate issue, however, is the highly insoluble nature of the acid; to circumvent this problem a known weight of acid was dissolved in a known volume of standard sodium hydroxide solution and back titrate the excess alkali.

0.2477g of the dried acid were dissolved in 25ml of 0.2025M sodium hydroxide solution. When solution was complete 2 drops of phenolphthalein indicator solution were added and the excess sodium hydroxide was titrated against 0.2M hydrochloric acid (note 1). 17.3ml were required indicating that 7.7ml were required to neutralise the acid:

7.7/1000 * 0.2025 = 0.0015593 M of acid - therefore 0.2575g of 4-amidobenzoic acid

Or 0.1295g of terephthalic acid (a dibasic acid so twice as much NaOH required to neutralise it)

Clearly therefore, the monobasic amidobenzoic acid looks the more likely candidate even if it is a little impure. It would appear from the greater than 100% content that the main impurity is likely to be incompletely hydrolysised PET.

For comparision the titration was repeated using the terephthalic acid recovered by the PET/caustic fusion process described previously.

0.2504 of terephthalic acid were dissolved in 25ml of 0.2025M sodium hydroxide. When the acid had dissolved 2 drops of phenolphthalein solution were added and the excess alkali titrated with 0.2M hydrochloric acid. 10.1 cc were added ie 14.9cc of sodium hydroxide solution were consumed neutralising the acid. Therefore:

14.9/1000 * 0.2025/2 = 0.001509 M of terephthalic acid = 0.2506 g ie almost exactly the amount taken, indicating that the terephthalic acid is eceedingly pure.

Further Work
When time permits I will dissolve this monobasic acid in the minimum amout of sodium hydroxide solution and attempt to recrystallise it to remove the partial hydrolysis products and isolate the pure acid. I also intend to attempt a faster process using aqueous ammonia in a pressure reactor that I now have to see if I can produce a purer product and a greater yield of the 4-acetamidobenzoic acid relative to terephthalamide.

Given the very low solubility of the terephthalamide in all the solvents I have tried makes further purification impractical so I am going to try and convert it into 1,4-dicyanobenzene and 1,4-diaminobenzene directly and purify these compounds.

Pumukli - 1-4-2016 at 07:04

Hats off, Boffis!

It seems that most of the ammonolysis product is the diamide. If you could find a decent route to the monoamide from the diamide it would be perfect! :-) I would (will) definitely join in when I have a bit better equipped lab. (I have rbfs now but still lack the ability to measure practically anything, melting point measurements included.)

Keep up and keep us informed!

Romix - 2-4-2016 at 01:14

Can sodium terephthalate be made by boiling crashed PET in NaOH.
Pan cooling down now. Boilled 200g of dust to 75g of base, 3 beakers full of water.
Some dissolved but not much, not even a gramm.

What about benzene from dry reagents? Is this true?

At room temperature, NaOH solutions some times etch PET, cracks it(Takes months to). Beware of leaks. HDPE are safe.
Isopropanol + NaOH made holes in PET bottle and cracked its neck in less then 10 min.


[Edited on 2-4-2016 by Romix]

Boffis - 2-4-2016 at 03:31

Hey can one of you moderators make my post above editable again. I have spotted a couple of omissions (note 1 is missing) and errors (the excess alkali was titrated against acid not alkali).

@Romix, to get rapid hydrolysis requires higher temperatures than you can get with dilute aqueous alkali solutions and after several experiments, some of which I described above, molten moist sodium hydroxide or a mixture of sodium hydroxide and a glycol type liquid seem to work best. The later method has the advantage of requiring a smaller excess of alkali and therefore acid to precipitate the product.

Your observations with alcoholic alkalis look interesting. Alcohol KOH is often used to hydrolyse compounds that hydrolyse slowly in aqueous NaOH so you might be on to something there.

[Edited on 2-4-2016 by Boffis]

Pumukli - 2-4-2016 at 10:59

Maybe the alcoholic solution can much better "wet" the otherwise hydrofobic PET surface than the simple aqueous alkali solutions?

Romix - 2-4-2016 at 19:37

Ok there's is to much water, I have to boil down to concentrate.
Boiled it one more time, at about 300C+ temp.
Pan were jumping and kicking, had to hold it with a glove, wearing gogles and dustmask.
Solution getting darker and darker.
Fumes coming of were sweet at one point, Wikipedia says that ethylene glycole is sweet tasting chemical.

[Edited on 3-4-2016 by Romix]

Romix - 2-4-2016 at 20:24

What's the solubility of sodium terephthalate? Any one knows?
If high, CO2 can be bubbled through filtered solution, and after filtering precipitate.
It can be recrystallized a few times in distilled water, then by adding acid, can be converted back to terephthalic acid?

Maybe you dear forum members know solvent for glue?
Recommend me some please.
On other forum one member told me that Xylene or Toluene can solve it.

In theory stickers should swim on top, plastic on bottom.
Cellulose cost more then carbon.
After wash solution in theory will separate in to layers.

[Edited on 3-4-2016 by Romix]

Boffis - 3-4-2016 at 17:19

Halogen in his post of 24-9-15 above suggested that the amide may contain short chain oligomers capped with ammonia (he means amide groups I presume). I think this is probably very likely but I couldn't see how to test it at the time. Having thought about it further and in the light of my results trying to test for amides with sodium hydroxide I think that one way to test it may be to heat some of the "amide" with strong (say 15-20%) aqueous ammonia to about 100 C in my newly acquired autoclave for 10-15 minutes. My idea is that the pure amide is rather resistant to further hydrolysis by ammonia but any uncleaved ester linkages will hydrolyse, some to amides and some to the ammonium salt, the latter are water soluble so the presence of water soluble ammonia salts in the aqueous phase is indicative of unhydrolysed ester linkages, ie there are residual oligomers.

@Romix, the solubility of normal sodium terephthalate is about 14g of salt per 100ml of saturated solution or about 15g of salt per 100ml of water based on my own experience and this is almost independant of temperature ie you can't recrystallise the sodium salt from water by simply cooling a hot saturated solution. Other sodium salts, particularly NaOH, greatly reduce its solubility so the salt can be precipitated by adding 60% NaOH solution until the concntration of NaOH in the terephthalate solution is about 15% but beware! The resulting slurry is impossible to filter with paper filters and is very fine grained and so washes through all but the finest glass frits. Both bases of terephthalic acid are stronger than carbonic acid and so CO2 will not precipitate terephthalic acid. Terephthalic acid is almost insoluble in water (about 10 milligrams per litre!), the only workable solvent for terephthalic acid appears to be dimethylformamide, solubility 7g per 100ml of solvent at room temperature.

I just peel off the labels from the bottles, this leaves small dabs of glue but these do not cause problems with such aggressive conditions so I don't try to remove them with expensive solvents.

I have also discovered a few more very basic transformation that PET will go through under modestly amateur friendly conditions: PET can be transesterified to dimethyl terephthalate with methanol and alkali (see US pat 3321510) and that this compound (solid Mp 142 C) can be half hydrolysed to methyl 4-amidobenzoate by alcoholic ammonia. According to another patent I found aqueous ammonia in an autoclave at 180-190 C causes complete hydrolysis to ammonium terephthalate.

Romix - 3-4-2016 at 20:30

Quote: Originally posted by Boffis  
Halogen in his post of 24-9-15 above suggested that the amide may contain short chain oligomers capped with ammonia (he means amide groups I presume). I think this is probably very likely but I couldn't see how to test it at the time. Having thought about it further and in the light of my results trying to test for amides with sodium hydroxide I think that one way to test it may be to heat some of the "amide" with strong (say 15-20%) aqueous ammonia to about 100 C in my newly acquired autoclave for 10-15 minutes. My idea is that the pure amide is rather resistant to further hydrolysis by ammonia but any uncleaved ester linkages will hydrolyse, some to amides and some to the ammonium salt, the latter are water soluble so the presence of water soluble ammonia salts in the aqueous phase is indicative of unhydrolysed ester linkages, ie there are residual oligomers.

@Romix, the solubility of normal sodium terephthalate is about 14g of salt per 100ml of saturated solution or about 15g of salt per 100ml of water based on my own experience and this is almost independant of temperature ie you can't recrystallise the sodium salt from water by simply cooling a hot saturated solution. Other sodium salts, particularly NaOH, greatly reduce its solubility so the salt can be precipitated by adding 60% NaOH solution until the concntration of NaOH in the terephthalate solution is about 15% but beware! The resulting slurry is impossible to filter with paper filters and is very fine grained and so washes through all but the finest glass frits. Both bases of terephthalic acid are stronger than carbonic acid and so CO2 will not precipitate terephthalic acid. Terephthalic acid is almost insoluble in water (about 10 milligrams per litre!), the only workable solvent for terephthalic acid appears to be dimethylformamide, solubility 7g per 100ml of solvent at room temperature.

I just peel off the labels from the bottles, this leaves small dabs of glue but these do not cause problems with such aggressive conditions so I don't try to remove them with expensive solvents.

I have also discovered a few more very basic transformation that PET will go through under modestly amateur friendly conditions: PET can be transesterified to dimethyl terephthalate with methanol and alkali (see US pat 3321510) and that this compound (solid Mp 142 C) can be half hydrolysed to methyl 4-amidobenzoate by alcoholic ammonia. According to another patent I found aqueous ammonia in an autoclave at 180-190 C causes complete hydrolysis to ammonium terephthalate.


Something precipitated on cooling, more like sodium carbonate.

[Edited on 4-4-2016 by Romix]

Romix - 4-4-2016 at 02:11

Boiled it down to 350ml. Something precipitating. White crystals.
Solution yellow. Glue might be the cause of colour change.
I let it sit for night. And then weigh a sample of precipitate and try decompose it at about 1000c+.
If it's therephtalate salt, there be decrease in weight. If not it's NaCO3.
Or simply be adding few drops of acid.

Romix - 4-4-2016 at 18:03

Crystals of IPA + NaOH that were difficult to scrape off.
Were dissolved in water. And PET bottle that were damaged by it.
Were cutten into squares and left in solution.

2 days sitting there. Noticeable cracks and holes every.

Ipa alone, doesn't react with PET.
Base does but very, very slow.



[Edited on 5-4-2016 by Romix]

Romix - 7-4-2016 at 05:51

Every time I check, there's more and more cracks and holes, on PET pieces soaked. IPA + NaOH= + H2O + PET
As what it goes into solution, any ideas dear colleagues?

Romix - 14-4-2016 at 12:14

Solution completly lost alchocholic smell. And turned yellow in colour after a week standing.
Wait few more months, see how it goes.

Sample that been boiled in sodium hydroxide, yellow too.

[Edited on 14-4-2016 by Romix]

Pumukli - 15-4-2016 at 02:55

Is the IPA smell lost? Strange. Because it has a strong, characteristic odor.

Although there is a process (trans-esterification) of vegetable oils catalysed by NaOH (or another strong base) in methanol what yields biodiesel, but you wrote that you added some extra water to the mix. And water is bad for that trans-esterification, it works against you (at least against the equilibrium). So I'm puzzled, I'm going to cut up some pet-bottles now and try to repeat this IPA-PET-Base "mystery". :-)

Romix - 15-4-2016 at 05:46

Quote: Originally posted by Pumukli  
Is the IPA smell lost? Strange. Because it has a strong, characteristic odor.

Although there is a process (trans-esterification) of vegetable oils catalysed by NaOH (or another strong base) in methanol what yields biodiesel, but you wrote that you added some extra water to the mix. And water is bad for that trans-esterification, it works against you (at least against the equilibrium). So I'm puzzled, I'm going to cut up some pet-bottles now and try to repeat this IPA-PET-Base "mystery". :-)


Yes.

Pumukli - 15-4-2016 at 07:58

I put 24.4g PET-slices and 2.4g NaOH (lentil-like half spheres) into a 250 ml FBF. Then I realized that all my chemistry equipment is in a transitional state in boxes and I could not locate the IPA bottle. :-)
Found some methanol so I used this instead of isopropanol. I poured 46 g methanol into that FBF and closed it.

The NaOH is still slowly dissolving in the flask, some of it is jammed between the PET slices above the methanol and has a really hard time dissolving. I think it will take a few days - if ever. :-) The solution is turbid by the way.
The flask is in my would-be laboratory, no heating in that building so the temperature is around 12 Celsius right now.

Pumukli - 16-4-2016 at 09:49

After 24 hours I can say that something is going on.

The methanol solution has small, white particles floating and is not transparent anymore. Not so white like e.g. milk yet but in another 24 hours it probably will look much like that. :-)

On the other hand there are no holes or cracks on my PET slices anywhere - which is sad, because I expected them. They have a thin, whitish (dust like) coating though - probably the same thing that is floating in the methanol.
I will take some pictures tomorrow and post them together with the yesterday made ones to show you the "before" and "after" states.

Maybe Romix has a different kind of PET with different additives over there (in the US?) and it may explain the cracks and holes he encountered during his experiment.
Or IPA is the real culprit, it maybe does something what methanol can not do under the same (khmm, at least very similar) conditions.

Pumukli - 17-4-2016 at 03:36

The original state, right after I closed the flask two days ago.

0hrs1.JPG - 51kB

Now, after the processes had 48 hours to run:

[Edited on 17-4-2016 by Pumukli]

48hrs1_1.JPG - 41kB

Another view from the bottom of the flask. You can see the white precipitate better from this angle:

[Edited on 17-4-2016 by Pumukli]

48hrs2_2.JPG - 52kB

There are no cracks or holes on the PET slices. The mixture does not seem to get yellow(ish) either, just forms that white precipitate. (The original PET bottle I cut up was a mineral water bottle, so the PET should be high quality (food grade), not recycled one. The slices originally had a light-blue hue.)

[Edited on 17-4-2016 by Pumukli]

Romix - 17-4-2016 at 15:59

Crystals of Methanol reacted with base?

Pumukli - 18-4-2016 at 02:46


Crystals of methanol?
My methanol does not have crystals, it is just the simple, free-flowing liquid. :-)

The white spots (crystals?) on the first picture are NaOH half-spheres. They dissolved very slowly in methanol at first giving a just slightly turbid solution. Since then that white precipitate formed. I think it is either terephtalic acid sodium salt or methyl ester of terephtalic acid.
I will check it later, in a few weeks, after the process has advanced and produced more of this compound (or consumed more of the original PET or NaOH, whichever happens sooner.)

Boffis - 18-4-2016 at 04:54

@Pumukli, thats interesting. It would be interesting to recover some of the white precipitate and see if it is water soluble or not. Sodium terephthalate is pretty soluble in water but dimethyl terephthalate is not and melts at 142 C if I remember correctly. This tell us whether sodium hydroxide in methanol causes transesterification or hydrolysis. If the methanol is dry I think you should get mainly the dimethyl ester which would be another useful starting material for synthesis.

Pumukli - 18-4-2016 at 06:24

I can't check the melting point (or just could provide a crude estimate of it) because I don't have a melting point apparatus (yet). But I will do my best. :-)

The water content of methanol is unknown, I just opened a (few) decades old bottle and used that. :-) It was originally "for synthesis" grade, I hope it is still in the "not too bad for amateurs" category.

I decidedly used 1/10th mass of NaOH to PET ratio because I was also curious whether it would transesterify or hydrolize the PET. I thought that 1/10th mass may mean "catalytic ammount", although it may be a bit overkill for that purpose due to the very different molar masses. Next time I will use less. :-)

Boffis - 18-4-2016 at 11:00

Actually Pumukli, if the precipitate melts easily and cleanly I think that's good evidence that its the methyl ester, if it melt with difficulty and decomposition then you have probably got the Na salt. I did some experiments with transesterification of phthalate esters some time back and it may be worth looking at:

http://www.sciencemadness.org/talk/viewthread.php?tid=23943#...

In this thread Nicodem posted an interesting reference in German that I have yet to try which is a sort of transesterification in the presence of urea to get phthalimide from phthalate esters. The idea has now occurred to me that this might be another route to terephthalamide via a similar route directly from PET.

I did some reading into transesterification in biodiesel production at that time and you are right that you only need trace amounts of strong base. This is the nice'n easy to follow home recipe, it formed the basis of my experiments in the referenced thread:


Attachment: Transesterification - Make Biodiesel (online) A M Helmensteine 2013.docx (68kB)
This file has been downloaded 284 times

Pumukli - 22-4-2016 at 01:53

Yesterday evening I filtered the precipitate from the PET-slices - methanol mixture becasue it was hard to see if it was still forming or the reaction ended.

Filtering was a pain.
I have only a glass fritt filter which fits my vacuum bottle and although the filtrate was crystal clear but the filtering took ages (well, more minutes than I expected or wanted to waste on this thing). I got a nice, voluminous white precipitate.

I checked its water solubility. Based on that what I got is not the transesterified PET but the hydrolized terephtalic acid sodium salt. :-( It dissolves in water fairly well and upon acidification with 2N HCl it immediately gives a dense, white precipitate. (Terephtalic acid)

Boffis - 24-4-2016 at 04:17

@Pumukli, it does sound like you got hydrolysis rather than transesterification. I wonder if it is possible to transesterify a polymer? I thought I had a paper specifically about transesterification of PET in methanol but i can't find it and I can't find any other applicable examples. There are several patents from various countries about a similar process but most seem to use high temperature and therefore pressures rather than a catalyst.

Pumukli - 24-4-2016 at 07:42

I agree, I could find only high temperature / high pressure methods for transesterification.

This low-temp method for hydrolysis is also not too good, I found a patent which uses fairly strong (approx. 90%) sulfuric acid for hydrolysis and according to that paper it takes around 10 minutes when PET is boiled in that acid. (Although it may take longer with pet slices instead of granules.)

Romix - 17-5-2016 at 16:36

What are decomposition products of Terephthalic acid?

Boffis - 19-5-2016 at 16:24

@Romix, can you be more specific. Decompose under what conditions? It is remarkably stable but thermal decomposition of the silver salt has been referenced above, I suggest you check out the reference. The whole essence of this thread is the conversion of terephthalic acid into other compounds, any of these reaction may be deemed a decomposition by some.

Eddygp - 20-5-2016 at 10:43

Has the product been identified as terephthalamide? As in, has anyone actually checked IR, NMR or at least melting point?

Boffis - 22-5-2016 at 05:51

Hi Eddygp, I don't have access to IR or NMR so no, The melting point is higher than I can measure (about 280 C). For various reasons I think that the terephthalamide is impure and contains a significant amount of short chain partial hydrolysis products. I intend to test this idea in the near future as part of my on-going work on terephthalic acid derivatives. I have observed that these partial hydrolysis products are suceptable to further hydrolysis giving a mixture of terephthalic acid and its monoamide and hope to be able to test the amount of each but I am still working on way to separate the two acids.

Eddygp - 23-5-2016 at 07:58

Quote: Originally posted by Boffis  
Hi Eddygp, I don't have access to IR or NMR so no, The melting point is higher than I can measure (about 280 C). For various reasons I think that the terephthalamide is impure and contains a significant amount of short chain partial hydrolysis products. I intend to test this idea in the near future as part of my on-going work on terephthalic acid derivatives. I have observed that these partial hydrolysis products are suceptable to further hydrolysis giving a mixture of terephthalic acid and its monoamide and hope to be able to test the amount of each but I am still working on way to separate the two acids.


Good. I'm trying to come up with an old school way to quantify the amount of terephthalic acid/monoamide there.
Maybe a reaction with the carboxylic acid taking advantage of its more reactive nature, to then quantify the amount of reagent that reacted away and give an estimate of the total number of carboxylic acid groups... oh well, this is when you realise how much merit the work of pre-IR/NMR/MS/... chemists has!

Boffis - 3-8-2016 at 13:38

Quote: Originally posted by Boffis  
Halogen in his post of 24-9-15 above suggested that the amide may contain short chain oligomers capped with ammonia (he means amide groups I presume). I think this is probably very likely but I couldn't see how to test it at the time. Having thought about it further and in the light of my results trying to test for amides with sodium hydroxide I think that one way to test it may be to heat some of the "amide" with strong (say 15-20%) aqueous ammonia to about 100 C in my newly acquired autoclave for 10-15 minutes. My idea is that the pure amide is rather resistant to further hydrolysis by ammonia but any uncleaved ester linkages will hydrolyse, some to amides and some to the ammonium salt, the latter are water soluble so the presence of water soluble ammonia salts in the aqueous phase is indicative of unhydrolysed ester linkages, ie there are residual oligomers.



I have now done some more experiments on the ammoniolysis products of PET plastic and it is fairly clear that hydrolysis is incomplete even in the alkali soluble fractions. I carried out the tests by reacting the various materials with 28% ammonia solution in my 50ml autoclave for an hour or so at about 100 C. The resulting solutions were acidified and the acid filtered off. A small, accurately weighed, sample of the dried acid was dissolved in a known volume of standard 0.2M sodium hydroxide (actually 0.2025M) and the excess alkali titrated against standard hydrochloric acid.

The white acid recovered from the large scale ammoniolysis experiment described above had a neutralization value that fell outside that of a mixture of terephthalic acid and its monoamide, almost certainly due to the presence of partial hydrolysis products. After "pressure cooking" with ammonia solution this value fell within the range for a mixture of terephthalic acid and its monoamide. In fact the results would appear to indicate that the new product is about 92% terephthalic acid monoamide and about 8% free terephthalic acid.

A similar experiment was carried out on the "terephthalamide" counterpart and after pressure cooking for an hour the ammonia soluble fraction was filtered off, acidified and filtered. The insoluble acid was subjected to analysis as described above and gave 83.7% terephthalic monoamide and 16.3% terephthalic acid.

The direct pressure cooking of new PET chips with excess ammonia solution under the same conditions resulted in only partial hydrolysis but the ammonia soluble portion gave 93.2% terephthalic monoamide and 6.8% free acid.

kmno4 - 26-10-2017 at 16:10

Nitration of TPA
Trial run.

To 100 cm3 Erlenmeyer flask with 30 g of conc. H2SO4 and magnetic dipole, 4 g of TPA is added, next 3,0 g of powdered KNO3 is added in portions. The white muddy mass is stired (or rather moved) by Nd-magnet (outside the flask) after each portion of KNO3. At the start, only 15 g of H2SO4 was used, but it gave not stirrable mass, so additional portion was added.
Next, the flask was placed on the water bath (exactly like in nitration of phtalic acid given in separate topic).
Heating at ~80-90 C during several hours gave nothing interesting - white mudd, slightly less dense on hot.
Heating was paused for 15 hours and continued the next day, but before heating, additional 0,5 g of KNO3 was added.
Soon after it was placed in hot water bath, I noticed the the white mass turn to bright yellow mass and became markedly less
dense. The clear yellow colour soon turn to brown-yellow one, but still bright. During another several hours of heating, the
colour did not change any more, but the mass started to increase its volume (foaming). The flask was removed from the bath and
set aside to cool down. The muddy mixture was less dense even at r.t., it may indicate that "something happened" (good or not good, hah).
The mass was transfered to 100cm3 of chilled water, giving intesively yellow solution with white foamy suspension, some
nitrogen oxide is generated but in so small amount that no hood is needed. It was set aside to cool down again and filtered*.
The white powder (I was almost sure it is unchanged TPA) in the funnel was washed with 2 x 10 cm3 of warm water and dried to
constant weight (4,1 g). I wanted to discadr it, but I noticed (during washing the equipment) the the powder is rather soluble
in water.
The powder was added to ~130 cm3 of water and heating was started. It dissolved gradually as temp. was increasing giving green solution. Several minutes of boiling caused complete dissolution and clear liquid, so it is not TPA !!!
On slow coolig, the solution deposits nothing during hours, but later small globules set on the bottom the flask, but in very small amouts. Keeping this for some time more at r.t. causes precipitation of white powder, supersaturation in this case is very stable, even if solid is already present.
Fitering gave 3,5 g of white** solid and pale green liquid which was kept for another run.

Another run :)
This time it was 15 g of H2SO4, 4 g of TPA, 3,1 g of KNO3. Procedure conducted as above, the white mud stirred from time to time with Nd-magnet (but I think it has no effect on the progress of reaction; no solid cake is formed during whole time of heating). First, 5-6 hours of heating, pause ~15 hours, and againg heating during ~8 hours. After ~6 hour of the second period, I noticed partial yellow coloration of the white mud and in a short time (<0,5 hour), the white colour turn to brown-yellow, as earlier. Heating was continued for additional ~2 hours (I had no mor time , it was late at night).
It was cooled, transferred into 75 cm3 of water. cooled and filtered. Obtained cake (~8 g), without washing and drying (it is
loosing the time and the product !) was dissolved on hot in green liquid from earlier run.
This time, some (however it looked like miligrams) solid was still present even on boiling, adding 10 cm3 of water gave nothing.
It was filtered on hot, green solution (but not completely clear) was quickly cooled down by immersing the flask in cold water.
Again, no solid was obserwed even cooled below r.t., solution became more turbid only (most possibly unreacted TPA). Suddenly, the solution turn milky and full of white precipitate.
Filtered, washed (~10 cm3 of ice cold water) and weighted: 4,7 g.
The most possibly it is desired nitro-TPA, unfortunately my melting point determination apparatus cannot be used (too low
temp.). I performed titration of this nitro-TPA, to estimate its purity: 0,44 g was titrated with 0,1 M NaOH. The result is 205
(as molar mass), it fits almost perfecly to 209 (theory).
The synthesis is dirty cheap, long-time heating is no problem, beacuse it can be paused at any time and continued later.
Of course, given proportions are surely not optimal but it works :)

* the filtrate deposits no solid on cooling (even below 0 C)
** it is purely white in LED-light, colour observed in day-light or incandescent lamp has faint yellow coloration

Boffis - 28-10-2017 at 04:58

@kmno4; magnificent! I am still away from home so I can't try this myself at present but when I get home I will have a go. This opens up a whole new avenue for derivatising terephthalic acid. It can be reduced to amine, acetylated and further nitrated and reduced to give the the dicarboxy-p-phenylenediamine. It may also be possible to replace the amine group via the diazonium salt with various groups. I wonder how reactive the bromo group would be in this position.

By the way, which way did you titrate the acid? I have found that for terephthalic acid it is better to use a back titration by dissolving the acid in a known volume of standard NaOH and then back titrating with standard HCl in the presence of phenolphthalein. This seems to give more accurate results.

Anybody got any info on pKa values for nitroterephthalic acid?

I am still trying to get my head round the paper you posted on the phase relationship of nitrophthalic acids and water.

The Volatile Chemist - 28-10-2017 at 07:57

KMnO4, did your NaOH have more sig figs than just 0.1M? If not, I hate to say it, but I don't think the titration shows much.

However, the reaction is sweet (something I've wanted to do) and the thread a sa whole is great!

[Edited on 10-28-2017 by The Volatile Chemist]

Boffis - 28-10-2017 at 11:49

Well the molecular weight of the terephthalate ion is 164 verses 209 for NTPa so that suggests about 91% NTP acid. however, in direct titrations against a slurry of terephthalic acid I found that the colour change occurs and then reverses and each time take a little longer to reverse. The pKa values for terephthalic acid are 3.51 and 4.82 so I would expect nitroterephthalic acid to have values more like 2.6 and 4.0 and I wondered if this might be a cause of some error. I have a nifty bit of software that save a lot of time and brain power on my personal computer unfortunately its a long way away!

The Volatile Chemist - 31-10-2017 at 10:30

Which piece of software? I'm curious, as I'm thinking of making a piece of lab-assistance software soon if I ever get free time myself.

I just meant that if you titrate with 0.1M NaOH, you have one sig fig to work with. Your MW, if it is calculated as 205g/mol, will be effectively anywhere between 150 and 249g/mol. Unless I'm reading something wrong.

Boffis - 1-11-2017 at 07:04

@Volatile Chemist, the software used here was BATE pH calculator. Its not free but I like it, its about 35 GBP for the download or you can buy the complete package of 4 pieces of software for about 120 GBP and that includes a formula balancing app that export into HTML so you can paste your equation into an SM post directly. I've been using the latter for years but only recently have I played with the other 3 pieces of software. You can download them for free and use them for a month to see if you like them. You can't save the charts in the free trial version but they are pretty interesting to play with.

I don't quite understand your second point. The NaOH titration gives you an equivalence number, which is the molecular weight per mole of ionisable hydrogen. So if you started with say 0.25g of acid and the titer was 33ml of 0.1M NaOH then 0.25g = (33 x 0.1)/1000 = 0.0033 moles if it's a monobasic acid or a MW of 75.76. If its a dibasic acid then there are only half as many moles present so 0.00165 mols = MW151.51.

When I do this type of titration on phthalic acids I dissolved about 0.2 g in 50ml of 0.1 or 0.2 M standardized NaOH solution and then titrate the excess NaOH against 0.1M HCl. I make up my own standard NaOH solution but I buy the acid a commercial standardized 1M HCl and dilute it to 0.1M and also use it to standardise my NaOH solution. So the accuracy of your MW determination is usually down to two main sources of error, one is the volume of titer which you can only read to an accuracy of about 0.05ml with my burette and the accuracy of the Molarity of the standard solutions but with my terephthalic acid work I was getting a MW of within 1% of the theoretical value, with the terephthalamidic acid there was a difference of 10+ % indicating that it is rather impure. So kmno4's 205 or MW of 207 indicates a mixture of about 91% nitroterephthalic acid 9% terephthalic acid assuming both are anhydrous and the titration is accurate.

kmno4 - 1-11-2017 at 12:41

Boffis, I made this titration with phenolphthalein as indicator, because it changes its colour at pH around 9, so it seems to be good choice. I think that purity you calculated (91%), is too low, but without HPLC I cannot prove it :)
I did reduction of this nitroacid with Fe. I chose this method because it is cheap and almost always works. However, I expected problems caused by the nature of reduced product.
With Fe powder, reduction is quick and takes 30 minutes to complete (with 4 g of nitroacid, 5 g of NaCl, 8 g of Fe, 75 g of H2O).
Unfortunately, prepared solution of aminoacid contains larger amounts Fe(II) - filtration gives clear solution, but after 1 minute it becomes turbid.
It is extremely hard to remove this iron from the solution (for me). I tried careful oxidation with 3% H2O2, but the solution gradualy turns red, most possibly some quinone is formed from the aminoacid. Additionally, formed Fe(III) hydroxide has very bad filtering properities: it comes through every filter I have.....
I have another reduced sample, without H2O2 treatment. It has dirty-pale-yellow colour, with reddish sedimemnt on the bottom.
Filtration gives clear pale greenesh solution, becoming turbid during almost seconds... I checked this solution for Fe(II) content with K3[Fe(CN)6] - the test is strongly positive. So far, I have no idea how to remove Fe from it in a simple way.
But it is really solution of desired aminoacid - it gives strong fluorescence (white-blue) under black-light lamp.

I am going to use Na2S2O4 as reductor the next time, but first I have to prepare some nitroacid (for now, nothing more left).
-------------
Addendum
The turbid solution, with Fe(II) contamination, has pH=5,5.
I was going to add some HCl(aq), but with bottle of this acid in my hand, voices in my head started to shout " Don't do this !".
Yeah, It was stupid idea, so instead of HCl I took NaOH solution and added some to the solution. The effect was terrific: suddenly whole solution turn blue-black for a moment, then some suspension started to float inside, after few seconds I got clear water-like solution with sediment looking like good old "Fe(OH)3".
It was filtered off (small amount) without problems, the filtrate was clear, with pale green-yellow colour, without tendency to become turbid in the air and pH=7,5.
Test for Fe(II) gave nothing, so damned Fe was removed completely.
The solution of amino-TPA is under further investigation.

ps.
0,1 M NaOH means what it means: 0,1 M NaOH
If you want to know more, the manufacturer says that concentration is between x mol/dm3 and y mol/dm3
But can I/you belive it ?
Discussion the difference between 0,1 and 0,14, accuracy of a balance made in PRC, selection of indicators for titration, pK of weak dibasic acid and its starting concentration.... is very interesting, but for separate topic(s).
I just wanted to know if I have TPA or nitro-TPA, so results 205, 210 or 199 are equally good estimations comparing to 166.
In general, performing accurate calculations for some substance with unknown impurities, is pointless.

[Edited on 2-11-2017 by kmno4]

The Volatile Chemist - 1-11-2017 at 15:45

Quote: Originally posted by Boffis  

I don't quite understand your second point. The NaOH titration gives you an equivalence number, which is the molecular weight per mole of ionisable hydrogen. So if you started with say 0.25g of acid and the titer was 33ml of 0.1M NaOH then 0.25g = (33 x 0.1)/1000 = 0.0033 moles if it's a monobasic acid or a MW of 75.76. If its a dibasic acid then there are only half as many moles present so 0.00165 mols = MW151.51.

When I do this type of titration on phthalic acids I dissolved about 0.2 g in 50ml of 0.1 or 0.2 M standardized NaOH solution and then titrate the excess NaOH against 0.1M HCl. I make up my own standard NaOH solution but I buy the acid a commercial standardized 1M HCl and dilute it to 0.1M and also use it to standardise my NaOH solution. So the accuracy of your MW determination is usually down to two main sources of error, one is the volume of titer which you can only read to an accuracy of about 0.05ml with my burette and the accuracy of the Molarity of the standard solutions but with my terephthalic acid work I was getting a MW of within 1% of the theoretical value, with the terephthalamidic acid there was a difference of 10+ % indicating that it is rather impure. So kmno4's 205 or MW of 207 indicates a mixture of about 91% nitroterephthalic acid 9% terephthalic acid assuming both are anhydrous and the titration is accurate.

Hmm...

Do you know what I mean by 'sig figs' or is this just a newer concept that not everyone has heard of? Nothing personal, just surprised - either you're missing something or I am.

If you buy an acid that says 1.M on it, it is likely pretty close to 1M. But how close?
To be able to write 1M on the bottle, it would need to be within the range of 0.5 to 1.4, or thereabouts. Now if the bottle said 1.0M, then we can rest assured it contains a solution between 0.95M and 1.04M.

So if someone wants to do a titration, and they use a solution they call 0.1M, then this is not a very accurate value to work with. For all we know, it could be 0.14M. However, if you refer to 1.00M NaOH, I know this solution is very trustworthy indeed. However, 0.1M, for a titration, is not trustworthy, and because we only have one digit of accuracy, the result of the titration can only have one digit of accuracy. And thus the molar mass calculated by it is not particularly reliable.

Boffis - 3-11-2017 at 07:14

I know exactly what you mean VC. What you are saying may be correct nomenclature in mathamatic but take a look at a bottle of commercial standard acid it says 1M, 2M, 0.2M etc not 1.00 or 2.00M. On a bottle of commercial 2M HCl that I purchase not so long ago it says in the small print that the acid is standardized to +/-0.5% of the stated value ie somwhere betewen 1.99 and 2.01 while on my Analar 1M acid it claims +/- 0.2% so somewhere between 0.998 and 1.002M. But I have no way of knowing the exact value so, like everybody else I use the stated value. I, like everybody else in the real world, only express more significant figure on solution that have been made by myself and compared against the reference solution, hydrochloric acid in this case.

So, please stop being a nit-picking-bastard and if you have nothing constructive to add to thus thread then don't post anything.

NEMO-Chemistry - 8-11-2017 at 08:48

TBH I think VC has a valid discussion, but i think this is completely the wrong thread. I found his argument thought provoking but it detracted from the main thread. So please start a thread on this, i would really enjoy something like that. But please not in this topic.

@kmn04

You made me go cut up milk bottles!! I got enough experiments running, i dont need/want another!! But its fascinating what your doing.

Up until now I confess to being in the 'meah' group of people when TPA is mentioned, to me its a white fluffy inert do nothing substance i have no use for.

But seems I got it all wrong lol, keep up the good work! I have my milk bottles in NaOH as we speak.

Boffis - 16-11-2017 at 09:39

@kmno4. I decided to try a "preparation scale" nitration of terephthalic acid based on your observations; here is what I did:-

20.045g of dry, reprecipitated, terephthalic acid were stirred into 70ml of conc. sulphuric acid (96-98%) in a 250ml flask containing a large stirrer bar. To this were added 15.40g of dried and finely ground sodium nitrate, initially in small portions but since there was almost no temperature rise eventually the remainder was added in one go. The flask was placed in a silicone oil bath on a stirrer hotplates and heated to 90 +/-3° C. The mixture proved difficult to stir so a further 15ml of conc. sulphuric acid were added. After about an hour a very slight reaction appeared to be occurring since, as kmno4 described, the mixture became thick and slightly frothy and a further 15ml of conc sulphuric acid were added to thin the reaction mixture. Once the target temperature had been reached the reaction slurry was left in the oil bath for 23 hours before being raised out of the oil and allowed to cool. The reaction mixture never really changed in appearance, it remained almost white and slightly frothy though the fine white suspension did appear to become more granular with time. The stirrer struggled to keep the frothy material stirred into the reaction mixture.

The cool reaction mixture was drown in 150ml of water and 150g of ice. A further roughly 100ml of iced water was used to rinse the flask. The result was exactly as you described, a bright yellow solution and a practically white, fine suspension that looks no different to terephthalic acid. Once the slurry had cooled to room temperature the solid was filtered off and washed with just a little iced water, about 30ml, and sucked as dry as possible before being air dried at 35-40°C. The yield of product was 23.334g of a fine white powder.

4.0635g of the crude product was recrystallised from 130ml of water as kmno4 described. It dissolved completely at close to boil point giving a clear pale yellow solution. It was left to crystallise overnight and cooled to about 8°C. The pure white product is still a fine white powder, which after filtering and drying gave 3.379g.

While I was waiting for the solution to crystallise I decided to analyse the crude product.

0.1774g of crude "nitroterephthalic acid" was accurately weighed out and dissolved in 20ml of 0.1865M sodium hydroxide (Note 1). One drop of phenolphthalein solution was added and the excess alkali titrated against 0.2M hydrochloric acid; 11.7ml were required:

Therefore 11.7ml of 0.2M HCl = 0.00234 Moles of HCl = 0.00234 Moles excess NaOH
NaOH used = 20.0ml of 0.1865M = 0.00373 Moles of total NaOH
Therefore NaOH used to neutralise the 2-ntp acid = 0.00373-0.00234 = 0.00139 Moles
It requires 2 moles of NaOH per mole of 2-ntp acid therefore acid = 0.000695 moles 2-ntp acid = 0.1774g

Therefore molar weight = 0.1774/0.000695 = 255.25

Errrrr that is way too high but is very close to a dinitro terephthalic acid but this is impossible given the ratio of nitrating agent to original acid.

Once the recrystallised material had dried it too was analysed in the same way using 0.1705g of this material.
This required 11.75ml of 0.2M HCl to neutralise the excess NaOH solution giving a MW of 247.1

This figure is lower but still too high. The calculated MW of various possibilities are:

2-nitroterephthalic acid = 211.13
any dinitroterephthalic acid = 256.13
nitroterephthalic acid monohydrate = 229.13
nitroterephthalic acid dihydrate = 247.16

This indicates that the product is most likely 2-nitroterephthalic acid dihydrate. Another possibility, though less likely given the dihydrate fit, is that some decarboxylation has occurred and the product is contaminated with 3-nitrobenzoic acid which would tend to increase the apparent MW because it is a monobasic acid.

@kmno4: The liquor left from the recrystallization is only pale lemon yellow and I wondered if you have some contamination of your terephthalic acid with the blue phthalocyanine pigment used to colour some plastic bottles? I only used clear, uncoloured bottles for my source of the acid.

Note 1 The standard NaOH solution was prepared from fine prills which tend to take up water from the atmosphere so although roughly 4g of NaOH were used to make up the 0.5 L of solution it still came out significantly below the target 0.2M. It was standardised against 0.2M HCl solution prepared from commercial Analar 2M HCl.

Late addition: The % yield of nitration product depends on what the product is. Since it now appears that the nitration product is a dihydrate that raised the question as to what the starting material was! So I analysed the raw terephthalic acid as I did for the compounds above and got a MW of 188.3 which is very close to the monohydrate which would be 184.15. Previously I had simply assumed that the terephthalic acid was anhydrous but appears that this is not so.

[Edited on 16-11-2017 by Boffis]

kmno4 - 19-11-2017 at 12:51

Nitration of TPA
Additional runs and interesting observations.

1) 6 g of TPA, 5 g of KNO3, 40 g of H2SO4 (95,5% by titration), total time of heating 23 hours at 75-80 C (total time of pause
about 37 hours). Even during this long-lasting heating, the "mud" inside the flask remains white. At the very end of heating,
the temperature of water bath was increased to 85 C and after short time, the reaction mixture started getting yellow.
So, this colour is rather some sign of decomposition, not the end of nitration.
2) Input as above, total mass of flask at the beginning of heating was 133,7 g, time of heating is 4+10 hours, with 6 hours
pause, temp. was kept between 70 and 80 C. The mass of the flask at the end of heating was 133,7 g. Cooled, white suspension was added to 100 g of ice-cold water (giving green solution+sediment), filter cake was washed with 2x3 g of cold water and dried. 7,1 g of white powder was obtained.
2a) 8,5 g of TPA, 57 g of H2SO4, 7,11 g of KNO3 (the mass of the flask before and after was 155,9 and 155,5 g), 15 hours of heating... 10,0 g of white powder was obtained.

The mass of combined samples of crude nitro-TPA is 24,0 g. 5 g of this acid was dissolved on hot in 100 g of water, chilled and
precipitated powder filtered. The green filtrate* was used for dissolving another 5 g portion... etc.
During some precipitation, the nitroacid deposited as crystalline powder, much easier to filter than the powder form and containing much less absorbed water.
About 0,5 g of this sample was kept for seeding subsequent solutions**.
Combined powders were combined and weighted: 21,8 g (dried). Titration of 0,32 g sample gave molar mass 214 (not bad).
It was purified in the manner as given above and interesting thing was observed: the hot solution of this partly purified nitroacid was practically colourless,ha.
So, the green coloration does not come from nitro-TPA, but from some highly soluble (fortunately) impurities.
About 21 g of crystalline powder (almost white) was obtained (looks similary to benzoic acid).
* these green solutions look the same as solution of K3[Fe(CN)6], at least for me
** solubility of nitroTPA seems to be 0,3-0,5 g per 100g of water (~10 C) and larger than 7 g (larger amounts were not tested) in boiling water.

Addition.
Boffis, your results and calculations seem to me a little aaa... incoherent. Your "hydrates" may exist only in your posts and the most possibly, it is the case. In this particular case of back titration, I would use methyl orange if your NaOH sol. contains larger amount of carbonate.
Next, I would strongly recommend drying your nitroTPA/TPA at least at 110 C (to constant weight). I have noticed that these acids, in powdery form, absorb rather large amounts of water when filtered off some solution. Vacuum filtration helps nothing, filter cake contains up to 50% water. It has nothing to do with hydrates, because the mass, after drying, does not change when kept in the air at r.t.
My experiments presented in this post were performed independently to yours, just simple coincidence. I have to repeat Fe reduction because it seems that I missed some important thing and the yield of aminoacid was close to zero :P

[Edited on 20-11-2017 by kmno4]

Boffis - 26-11-2017 at 12:50

@ kmno4, I been reading up about terephthalic acid and its nitration product again and there is no mention of them forming a hydrates. This is a bit of a mystery to me. Since nitroterephthalic acid is fairly soluble it is possible to titrate it directly so when I am home next I will try this out. I have also ordered some new standard 2M HCl from a reputable supplier just in case my old 2M HCl is not accurate any more (I can't even remember where it came from but I have had it for many years). I will also check the melting point of the nitroterephthalic acid which is significantly lower than the parent acid.

By the way I make up my standard sodium hydroxide solution by dilute a 60% (15M) stock solution because the sodium carbonate formed by absorption of CO2 from the atmosphere is virtually insoluble in strong NaOH and forms a sediment which can be avoided by decanting carefully and diluting. I then standardize it against my 2M HCl (usually dilute to 0.1 or 0.2M). However the last thing I did before leaving home last week was to try titrating my diluted HCl against a primary standard solution made of BDH Analar sodium tetaborate and I did not get good results so it looks like my primary standard my be the problem; hence the new acid.

unionised - 26-11-2017 at 14:59

As a slight aside. This looks like a good candidate for a back titration.


[Edited on 26-11-17 by unionised]

Boffis - 26-11-2017 at 23:29

Quote: Originally posted by unionised  
As a slight aside. This looks like a good candidate for a back titration.


[Edited on 26-11-17 by unionised]


What does?

I have discussed my use of back titration above on terephthalic acid due to its low solubility. The nitroterephthalic acid is sufficiently soluble to dissolve 0.2g in 25ml of water (see kmno4's comments above) so I think that it is a candidate for normal titration. Don't you?

kmno4 - 29-11-2017 at 02:34

Reduction of nitro-TPA (II)
a) Fe
2 g of nitro-TPA, 2 g of NaCl, 2 g of pulverized Fe, ~50 g of water.
The acid is dissolved in the water on hot, NaCl is added. To this solution, the iron is added in one portion.
The solution turns green during seconds, then dirty-green -> more dirty -> dirty-green-brown -> dirty-brown, all the time
sediment is produced. As soon as Fe is added, mechanical stirring must be applied to prevent creation of Fe/Fe oxides chunk,
hard and damned strongly adhering to glass. Stirring with magnetic dipole is not good solution, because its plastic coat will
be damaged by hard particles.During first ~30 minutes, the mixture is hot by self generating heat, after this time it is
externally heated to 70-90 C (mixing with glass rod at minutes intervals is applied) during additional 1,5 h (or longer)*.
After given time, the mixture is cooled and calculated (as for disodium salt + 10-20% more) amount of NaOH is added.
This gives dirty-yellow-brown liquid and brownish sludge. This is carefully filtered, giving nicely looking water-clear yellow
solution and Fe oxides cake. If the solution is reddish/pinkish ... etc, possibly filtering was not perfect and very tiny
amount of Fe oxides are present.
Calculated amount (based on NaOH added) of HCl(aq) is added - this causes precipitation of yellow, very tiny powder, sometimes
semi-colloidal: it is desired amino-TPA. The mixture it then heated to weak boiling, this operation gives better filterable
sediment. The yellow powder is then filtered, washed and dried.Assuming it is pure amino-TPA and pure nitro-TPA was used, the yield is about 90% of theory. Pinhead amount of this powder, dissolved in ~20 cm3 of alkaline water, gives strong, nice fluorescence under UV light.
The strength of the fluorescence depends on amino-TPA concentration and pH. The best results are obtained in highly diluted, weakly alkaline solutions. The yellow amine is also fluorescent (orange) in solid state.
* heated for shorter time and/or bad stirring, may give low yield (>50%) of amino-TPA. However, unreacted nitro-TPA solution can be reduced again with Fe.

b) Na2S2O4
Seems to be optimal choice for the reduction.
Unfortunately - I was not able to obtain more than 40% yield of amino-TPA. Nitro-TPA is reduced completely, because additional
amount of dithionite gives nothing. First portions of NaS2O4 sol. gives interesting deeply-red solution, but it turns yellow,
when about 2/3 of calculated amount* is added. I have no idea why I get such low yield.
* 3 moles of Na2S2O4 per 1 mole of nitro-TPA

[Edited on 29-11-2017 by kmno4]

Boffis - 29-11-2017 at 08:54

@kmno4. I wonder if for the dithionite reduction it would be better to dissolve the Nitro-TPA in just sufficient NaOH to give the neutral salt first as dithionite is unstable in acid conditions. The red colour is probably a transient azo compound formed by the condensation of an intermediate nitroso compound with resulting amine. I got a similar result reducing nitro-phenylpyrazolones but here the compound is so insoluble I was able to isolate the azo compound. Further reduction converts it into a hydrazo-derivative but some of these are fairly resistant to further reduction. It might be better to add a sodium salt of N-tpa solution to the dithionite solution to try and limit the amount of transient nitroso that is present at any time.

kmno4 - 29-11-2017 at 12:38

For Na2S2O4 reduction it was:
2 g of nitro-TPA, 5,8 g of Na2S2O4 (assuming ~85% purity), 2,3 g of NaOH.The acid was dissolved in water with 2 g of NaOH, dithionite was dissolved in ~20 cm3 of water containing the rest of NaoH (0,3 g). This solution was added, in parts, to the solution of just prepared Na2nitroTPA salt. Results as above: red solution, gradually turning yellow, yield of the amine after careful acidification (HCl) around 40%.
This was tried again, with slightly different proportion, but with the same results. Extremely fine suspension of amine powder, looking like about 5 g or so, gives after filtering and drying dissapointing amount of less than 0,7 g of fine yellow powder (it is not sulfur, hah). The filtrate remains yellow, but I cannot force it to give me another portions of amine. Swinging pH gives nothing, but SO2 formation and its dissapiring.

kmno4 - 16-12-2017 at 14:18

Quote:
For Na2S2O4 reduction it was:
2 g of nitro-TPA, 5,8 g of Na2S2O4 (assuming ~85% purity), 2,3 g of NaOH.The acid was dissolved in water with 2 g of NaOH, dithionite was dissolved in ~20 cm3 of water containing the rest of NaoH (0,3 g). This solution was added, in parts, to the solution of just prepared Na2nitroTPA salt. Results as above: red solution, gradually turning yellow, yield of the amine after careful acidification (HCl) around 40%.(...)

The yield is increased to slightly above 90%, but rather large amount is needed. It seems that Na2SO3, NaHSO3 and SO2 gives some kind of buffer and really large excess of acid is needed to liberate the amino-TPA.
In this way, my experiments with TPA are finished ;)

Boffis - 17-12-2017 at 10:22

Hi kmno4, to get the 90% yield do you mean you had to add a lot more hydrochloric acid? I'll be home in a few days so I'll have a go at this.


kmno4 - 20-12-2017 at 02:51

Quote: Originally posted by Boffis  
Hi kmno4, to get the 90% yield do you mean you had to add a lot more hydrochloric acid? I'll be home in a few days so I'll have a go at this.

Hi, Boffis
Yes, just a typo, I mean HCl(aq).
Calculations were made according to this:
TPA-NO2 + Na2S2O4 + NaOH -> TPA-NH2 + Na2SO3 + H2O
It seems that obtained solution is too basic or TPA-NH2 gives some salt/adduct with SO3/HSO3 and additional portion HC(aq) is added to liberate almost all SO2. Besides, on 4,5 g TPA-NO2 scale and ~200 cm3 of solution, after some time, on cold, large amount of some crystals are formed, possibly Na2SO3 hydrates.
However, when too much hydrochloric acid is added, TPA-NH2 dissolves again and some sulfur slowly precipitates, possibly from
Na2S2O3 formed by the way, but it can be reversed by adding some NaOH.

Boffis - 12-10-2019 at 11:16

I have recently returned to the issue of terephthalic acid from PET and have discovered that my terephthalic acid is actually rather less pure that I had imagined. The main impurity appears to be short chain glycol terephthalate polymer units with one or two free acid units at the end. They are very difficult to separate from the free terephthalic acid but I am going to try some experiments with secondary hydrolysis carried out in much the same way as for the original PET but carried out on the crude sodium terephthalate.

The problem came to light after the discussions with kmno4 on the nitration of terephthalic acid where I couldn't get an accurate neutralisation equivalence for the acid and thought that the free acid may be a hydrate. I have now found that it can be purified by converting the nitroterephthalic acid to the mono sodium salt by dissolving in a small amount of hot water (2.5ml per gram) containing 0.9 molar equivalences of sodium hydroxide. The milky white slurry is then treated with charcoal and and filtered hot and cool. If the sodium salt starts to crystallise as white granules the solution should be reheated and threated with charcoal again. When the solution is fairly pure the sodium hydrogen terephthalate starts to crystallise as delicate colourless needles and eventually the whole solution is immobilised by a network of such needles. These should be filtered off and recrystallised from fresh water (2ml per gram). If the solution is completely clear its OK if not it should be treated with charcoal again.

The filtrate from the first crop of crystals can be similarly worked up by evaporating down to a third of its orginal volume, charcoal treating and cooling.

The free, pure nitroterephthalic acid can be recovered by acidifying the hot slightly diluted solution and cooling.

Both terephthalic acid and its nitro compound for di-sodium salts that have almost flat solubility curves so they cannot be recrystallised by cooling a hot concentrated solution. This acid gives a nearly theoretical neutralisation value.

Fulmen - 16-10-2019 at 11:22

Question: Is there any risk of getting amides from TPA and aqueous ammonia? I'm wondering if boiling a solution of ammonium terephthalate would drive off the ammonia, allowing a more dense precipitate to form.

Fulmen - 19-10-2019 at 11:15

Good news everybody. This trick worked pretty OK. Starting with 56g of PET I obtained almost 50g of crude TPA that could be filtered and dried into a free-flowing powder. That's a bit more than 100% yield, I expect it to drop a bit after a second wash.

I did have a lot of trouble with bumping, as I didn't have any suitable boiling stones I simply separated the ppt by decanting at regular intervals. But other than that it seems to work.

Boffis - 19-10-2019 at 13:43

@Fulmen; I am not sure what you are getting at but I don't think that evaporating terphthalic acid solution in ammonia will generate the diamide of terephthalic acid, I doubt that it will generate even much mono-amide but I think that terephthalic acid is probably strong enough to hang on to most of its ammonia. Its an interesting idea though and worthy of a bit of experimentation.

I have done some further work on the mono sodium nitroterephthalate and found that it crystallises as the dihydrate.

Attempts to purify the terephthalic acid by the same process failed, only half the acid dissolved forming the normal salt. I can't believe all of the remaining material is impurities though. I converted 90g of my crude terephthalic acid into the normal sodium salt and then fused it with a mixture of ethylene glycol, sodium hydroxide prills and a little water for three hours. The resulting porridge like mass was diluted with 100ml of water and filtered through glass. I will dissolve the mass in water in the morning and acidify it to precipitate the free acid and see if I can get better results.

 Pages:  1