GB1604930A

GB1604930A - Process for the synthesis of terephthalic acid

Info

Publication number: GB1604930A
Application number: GB25950/78A
Authority: GB
Original assignee: Montedison SpA
Current assignee: Montedison SpA
Priority date: 1977-07-26
Filing date: 1978-05-31
Publication date: 1981-12-16
Also published as: ES470322A2; DE2823504C2; NL7805739A; JPS5439038A; IT1080758B; CA1113957A; FR2398716A2; FR2398716B2; DE2823504A1

Description

(54) PROCESS FOR THE SYNTHESIS OF TEREPHTHALIC ACID (71) We, MONTEDISON. S.p.A., an Italian body corporate, of 31 Foro Buonaparte, Milan, Italy, do hereby declare the invention, for which we pray that a patent may be granted to us, and the method by which it is to be performed to be particularly described in and by the following statement: The invention relates to an improvement in a process for the synthesis of terephthalic acid comprising the oxidation of paraxylene in acetic acid solution, according to the equation:

in the presence of a catalytic system containing manganese, bromine and cobalt.

Japanese Patent Specification No. 66643/1974 discloses that manganese and cobalt acetates are suitable raw materials for the preparation of the catalytic system, that bromine can be added as hydrogen bromide or elemental (free) bromine (but never as alkaline bromide) and that acetaldehyde can be advantageously present in the oxidation zone. This disclosure is not entirely correct and the teachings are only applicable when bromine is added as hydrogen bromide. Should in fact the catalytic system be prepared in acetic acid solution from manganese acetate (or metallic manganese) and elemental bromine, (which preparation has neither been exemplified by the Japanese publication above nor by other patents according to our knowledge) there is a strong danger of manganese dioxide precipitation (MnO2), which would cause the clogging in the apparatus. The plants involved in such processes are very big and costly and such clogging would cause enormous losses of output capacity. This is particularly true when the water content is higher than 1% by weight with respect to the acetic acid and when the bromine: manganese weight ratio is in excess of 1.65. Sometimes the presence of as little as 0.2% by weight of water can cause very troublesome clogging and consequently great output losses.

The use of elemental bromine (free bromine) could have many advantages over the use of hydrobromic acid. Bromine is readily available as it comes directly from natural sources, while hydrogen bromide is prepared from bromine through a series of troublesome chemical stages. Furthermore hydrogen bromide is a gas which is commonly sold as nearly 50% solution having a density of about 1.5 g/cm3, while the density of bromine is about 3 g/cm3, therefore the storage volume for hydrogen bromide solutions must be four times the corresponding storage volume for bromine.

It is an object of the invention to provide a process for the bromine catalyzed synthesis of terephthalic acid in which the source of bromine is elemental bromine.

According to the invention there is provided a process for the synthesis of terephthalic acid, comprising oxidizing paraxylene in acetic acid solution in the presence of a catalytic system containing manganese, bromine and cobalt, in which bromine is added as elemental bromine, separating the synthesized solid terephthalic acid from the mother liquor and rectifying at least a portion of the mother liquor to reduce the water content therein, in which: a) the catalytic system is prepared by bringing into contact acetic acid, manganese or manganese compounds, elemental bromine, a reducing compound, water and a soluble cobalt compound, b) the catalytic system obtained according to a) and containing manganese, bromine and cobalt in a dissolved form, is fed to the oxidation zone, optionally in admixture with the paraxylene feed and/or with at least a portion of the rectified mother liquor.

In a preferred embodiment, the manganese is present in an amount of from 50 to 1000 mg/Kg of acetic acid, the manganese:cobalt weight ratio is from 2:1 to 4:1 and the bromine:manganese weight ratio is preferably a maximum of 1.65:1 and generally from 0.5:1 to 1.65:1, the amount of water is preferably a maximum 1%, more preferably a maximum of 0.2% by weight, with respect to acetic acid, but higher amounts can be present, provided there is a reducing compound:bromine molar ratio of at least 1 preferably from 1 to 2, and more preferably from 1.1 to 1.5. Acetaldehyde e.g. reacts according to the following equation:

but minor amounts of bromo-organic compounds are also formed, e.g. monobromo-acetic acid and monobromo-acetaldehyde. These compounds release the hydrobromic acid, slowly by hydrolysis.

The preparation of the catalytic system used in the invention is slightly exothermic at the start; but in a short time supplemental heating is needed and it is advisable to operate between room temperature and the boiling point of acetic acid, e.g. from 20 to 1200C, preferably from 40 to 90"C. Cobalt is preferably added as cobalt carbonate or acetate (commerically available in the tetrahydrate form) and manganese can be added as finely divided metal or in the form of carbonate or acetate. Other elements, e.g. chromium, can be optionally included.

Another simple and suitable source of manganese and cobalt is the recovery mixture coming from the treatments of the purge of the terephthalic acid synthesis, i.e. the purge of the synthesis mother liquor. Such recovery mixture usually contains acetates or carbonates and strongly oxidised compounds of manganese which are not easily dissolved. The use of acetic acid-acetaldehyde solutions will quantitatively attack every form of recovered cobalt and manganese and promote the conversion of bromine to hydrobromic ion.

The catalyst system may be prepared by initially bringing into contact acetic acid, manganese and elemental bromine, in the presence of a reducing compound, preferably acetaldehyde, and then adding the soluble cobalt compound. Besides acetaldehyde other aldehydes can be used, e.g. paraldehyde, isobutyraldehyde, paratoluic aldehyde, paracarboxy-benzaldehyde and mixtures thereof. Other suitable reducing agents are the aliphatic alcohols preferably those having from 1 to 5 carbon atoms, e.g. ethanol and isobutyl alcohol, diethyl-acetal and ethylidene-diacetate. Paraxylene itself can be considered a reducing agent.

The invention will now be illustrated by the following Examples and the accompanying drawings in which Figures 1 and 2 represent diagrams of plants suitable for the preparation of terephthalic acid.

Examples 1 to 8 are not in accordance with the invention and show that the presence of a reducing agent can be dispensed with under certain precise conditions.

Example I 32.3 parts by weight of electrolytic manganese in the form of flakes having a thickness of from 1 to 2 mm and a surface of from 0.5 to 5 cm2, 150 parts by weight of acetic acid containing 0.2% by weight of water, and 46 parts by weight of liquid bromine at 99.8% by weight were loaded singly in that order into a stainless steel autoclave, equipped with a rotating stirrer, a relfux cooler and a heating jacket. The resulting bromine:manganese ratio was 1.42 by weight. A slightly exothermic reaction began which raised the temperature to 450C and heat was then supplied to maintain the temperature at such value throughout the reaction. After all the bromine had reacted, a suspension consisting of residual manganese bromide and manganese acetate in acetic acid was obtained; under these conditions, the reaction of manganese with bromine occurred more quickly than the reaction with acetic acid. 231 parts by weight of water were added, to dissolve both the bromide and acetate and promote the attack of residual manganese by acetic acid. The resulting solution had the following composition: Manganese (Mn++) 7.03% by weight Bromine (Br-) 10.00% by weight Free acetic acid 24.82% by weight Water 50.20% by weight.

Tetrahydrated cobalt acetate was added to this solution in an amount as to obtain a Mn:Co weight ratio of 4:1 and a Br-:(Mn++ + Co++) weight ratio of 1:4. The resulting solution was used as catalyst in the oxidation of paraxylene to terephthalic acid and the results obtained were excellent.

Examples 2 to 4 Example 1 was repeated varying the amount of cobalt in order to provide Mn:Co weight ratios respectively of 1:1, 2:1 and 3:1, corresponding to Buy : (Mn++ + Co++) weight ratios of 0.71, 0.95, 1.06. The results obtained were similar to those of Example 1.

Examples 5 to 8 Example 1 was repeated raising the weight ratio of Br to Mn from 1.42 to 1.65 and analogous results were obtained. The ratios of Br- to (Mn++ + Co++) were respectively 1.32, 0.82, 1.10 and 1.24.

Example 9 The following products were introduced into the autoclave used in Example 1: 19 parts by weight of electrolytic manganese in flakes, 95 parts by weight of acetic acid at 99.8% (by weight), and 41.8 parts by weight of liquid bromine at 99.8% (by weight). The Br:Mn weight ratio was 2.2. The temperature spontaneously rose to 320C and was further increased to 450C by supplementary heating. This temperature was maintained for 8 hours, to obtain an almost total conversion of bromine (98.21%). The addition of water would cause, in the presence of a strong oxidizing agent such as unconverted bromine, the precipitation of MnO2 and therefore the risk of clogging the lines. This risk can be eliminated by removing the precipitate, although this requires complex operations and large apparatus.

According to the process of the invention 0.275 parts by weight of acetaldehyde were added per 1 part of unreacted residual bromine, whereby the bromine was immediately converted to hydrobromic ion. Water was then added (304 parts by weight) and the reaction of manganese with the acetic acid terminated. Tetrahydrated cobalt acetate was then added in an amount to provide a Mn:Co weight ratio of 4:1 and a Br-:(Co++ + Mn++) weight ratio of 1.76. There were no cloggings or other problems of any kind during the preparation of the catalytic solution and during its utilization in the oxidation of paraxylene to terephthalic acid.

Examples 10 to 12 Example 9 was repeated, varying the amount of cobalt acetate, to provide Mn:Co weight ratios of 1:1, 2:1 and 3:1, corresponding to ratios of Br- to (Mn++ + Co++) of 1.1, 1.46 and 1.65. The results were very satisfactory, especially when the ratios 2:1 and 3:1 were used.

Example 13 Run a) 14.25 parts by weight of electrolytic manganese in flakes, 62.60 parts by weight of acid at 99.8% by weight, and 143.10 parts by weight of water were charged successively in that order into the autoclave used in Example 1. The temperature was maintained at 650C for 8 hours under stirring, until all the manganese was converted to the corresponding acetate.

The hydrogen evolved was diluted with nitrogen and vented to the atmosphere. To the resulting solution, containing about 14.5% of free acetic acid, 6.91 parts by weight of acetaldehyde were added, and 20.9 parts by weight of liquid bromine (Br:Mn = 1.46) were gradually fed over a period of 45 minutes. The bromine was completely reduced without any oxidation of manganese.

Run b) Employing the same conditions as in Run a) above, the catalyst system was prepared by feeding acetaldehyde in the form of diethyl-acetal (18.53 parts by weight) and the bromine was gradually added over a period of 45 minutes and quantitatively reduced without precipitation of MnO2. 30.08 parts by weight of tetrahydrated cobalt acetate were then added, to yield a concentrated catalytic solution suitable for use in the paraxylene oxidation reactors.

The Mn++/Co++ weight ratio was 3:1 and the Br~/(Mn++ + Co++) ratio was 1.

Run c) The procedure of Run a) above was repeated employing ethylidene diacetate in an equivalent amount instead of acetaldehyde g22.92 parts by weight of ethylidene diacetate).

The reduction of elemental bromine was quantitative. The final, perfectly limpid and concentrated catalytic solution, having a ratio Mn++/Co++ of 3:1 and a ratio Br/(Mn++ + Coy+) of 1, was then conveyed to the paraxylene oxidation reactors.

Example 14 Example 9 was repeated, but raising the temperature to 800C and obtaining a full conversion of the reagents.

Example 15 This Example utilizes the apparatus illustrated in Figure 1. The mother liquor coming from the centrifugation of terephthalic acid (at 100 to 1300C), containing about 90% by weight of acetic acid, was fed, through line (1), above the third tray from the bottom of a rectifying column (stripper), to the head of which was fed through line (2), a stream of aqueous acetic acid at about 70% by weight, free from catalyst, formed by condensation of the vapours released during to the oxidation of paraxylene. The acetic acid:paraxylene weight ratio during the reaction was 3:1 and the water contained in the reaction mixture was between 3 and 4 % by weight. In the fresh mixture (14) which was fed to reactors the acetic acid:paraxylene weight ratio was about 4:1, the value of this ratio decreased successively, because of the extraction of a portion of the condensed vapours at the synthesis temperature.

The bottom liquid of the stripper, which operated at a temperature of about 1300C or less, contained about 0.03% by weight of suspended solids consisting predominantly of terephthalic acid which is preferably recovered. A portion of the bottom liquid, containing about 97% by weight of acetic acid and about 50% of the catalyst fed to the column, together with oxidation intermediates, was recycled, through line (3), to the tank containing the feed mixture for the oxidation reactors. A second portion of the bottom liquid passed to the column reboiler, consisting of a still pot, equipped with a stirrer, and of a heat exchanger, arranged in series and steam-heated. A portion of the concentrate leaving the pot formed the process purge and was sent, through line (4), to a thin layer evaporator (not shown in the Figure) and then to an incinerator. The remaining portion of the liquid leaving the tank passed through the exchanger and flowed back to the still pot, where the vapours were released and flowed back to the column bottom. The residence time on the column bottom, about 6 minutes, was not so long as to alter the organic substances of the recycled solution (3).

The vapours (15) passed from the column head to a second (azeotropic) column for the recovery of acetic acid, equipped with a reboiler, a reflux condenser and a mixing tank; into which tank, an amount of azeotropic agent (isobutyl acetate) sufficient to make up for the losses was added through line (5). The organic phase that separated, flowed back to the column, while the aqueous phase that collected in a trap passed, through line (6), to a stripping column for the recovery of the azeotroping agent and of the methyl acetate present therein. The acetic acid flowing out from the bottom contained 3% by weight of H2O and was passed through line (7), to the tank containing the feed mixture, where it was admixed with the recycle coming from line (3) and with fresh paraxylene coming from line (8).

The catalyst system was prepared in a stainless steel autoclave, equipped with a rotating stirrer, a relfux condenser and a heating jacket. The following components were successively added 14.25 parts by weight of electrolytic manganeses in flakes through pipe (10), 62.60 parts of 99.8% acetic acid by weight, through line (9) and 143.10 parts by weight of water through line (12).

The temperature was maintained at 65"C for 8 hours under stirring until all the manganese was converted to the corresponding acetate and the hydrogen evolved was diluted with nitrogen and vented to the atmosphere. To the resulting solution. containing about 14.5% of free acetic acid, 6.91 parts by weight of acetaldehyde were added, and 20.9 parts by weight of elemental bromine (Br:Mn = 1.46 by weight) were gradually fed over a period of 45 minutes and the bromine was completely converted to hydrobromic ion without any precipitation of MnO2. 30.08 parts by weight of tetrahydrated cobalt acetate were then added, to obtain a concentrated solution that was proportioned into the tank where the feed mixture was prepared. The Mn:Co weight ratio was 3:1 and the Br-: (Mn + Co++) weight ratio was 1:1, this ratio of the fresh concentrated catalyst took into account the bromine losses occurring during the reaction.

Example 16 89.3 parts by weight of a mixture of manganese carbonate and of cobalt carbonate (equal to 32.3 parts by weight of manganese and 10.76 parts by weight of cobalt) mixture recovered from the ashes of the synthesis purges coming from the mother liquor treatments, were charged into the autoclave through line (12) in Figure 1. 212 parts by weight of acetic acid (containing 12.9 parts by weight of acetaldehyde) and 367 parts by weight of water were added through line (9) thereto. The temperature was raised to 72"C and 46 parts by weight of liquid bromine were fed through line (11) by means of a submersed plunger, over a period of 40 minutes, under stirring. The mixture was further stirred for several hours at 72"C to yield a perfectly clear solution containing acetates and bromides of manganese and cobalt in the most suitable ratios for the terphthalic acid synthesis.

Examples 17 to 26 Example 15 repeated, but varying the amount of the reagents and raising the temperature to 78"C. At the conclusion of the reaction residual bromine could never be found, with the exception of Example 26, where the presence of residual bromine in an amount of 0.01% by weight was ascertained.

The results obtained are recorded in the following Table I.

TABLE I Example Molar ratio Percentages by weight Time (minutes) CH3CHO:Br2 Feed Final Product Charge Reaction Co++ Mn++ H2O Br2 Br17 4.2 - - 4.8 0.17 0.10 - 15 18 4.1 0.037 0.117 5.3 0.17 0.15 - 10 19 3.8 0.037 0.117 5.2 0.159 0.15 20 15 20 2.6 0.037 0.116 14.4 0.16 0.16 20 30 21 3.4 0.037 0.116 29.7 0.16 0.16 20 25 22 3.7 0.04 0.12 30.2 0.14 (*) 18 21 23 3.4 0.037 0.114 49.95 0.16 0.15 20 0 24 2.9 0.036 0.113 69.2 0.15 0.14 20 0 25 1.76 0.85 2.90 71.9 7.91 (*) 20 0 26 1.1 0.77 2.35 33.3 3.83 2.42 20 8 (*) Not measured Examples 27 to 37 Operating according to Examples 17 to 26 tests were carried out using other aldehydes in place of acetaldehyde. The results are recorded on Table II.

TABLE II Example Aldehyde Molar Percentage by weight Time (minutes) ratio aldehyde: Feed Final Br2 Product Charge Reaction Co++ Mn++ H2O Br2 Br27 Paraldehyde 1 - - 5.0 0.17 0.11 - 15 28 Paraldehyde 0.95 0.039 0.12 4.99 0.178 0.16 - 75 29 Paraldehyde 4 0.04 0.12 4.99 0.16 0.14 - 45 30 Paraldehyde 1.3 - - 5.2 0.13 (*) 18 19 31 Toluic Aldehyde 3.2 - - 5.02 0.178 0.13 - 400 32 Toluic aldehyde 3.3 0.019 0.06 5.02 0.08 0.066 - 1100 + 95 33 Toluic aldehyde 3.4 0.039 0.12 5.0 0.157 0.14 - 75 + 105 34 Toluic aldehyde 2.11 0.037 0.117 5.2 0.154 (*) 20 270 35 Isobutyric aldehyde 2.7 0.037 0.116 50.2 0.159 0.16 20 120 36 Isobutyric aldehyde 3.6 0.34 0.113 68.11 0.14 0.14 40 120 37 Isobutyric aldehyde 1 0.74 2.28 33.12 0.390 3.54 20 120 Example 38 This Example utilizes the apparatus illustrated in Figure 2. The mother liquor coming from the centrifugation of terephthalic acid and containing about 85% by weight of acetic acid, was fed, through line (1) above the third tray from the bottom of a rectifying column, to the head of which a stream of aqueous acetic acid at about 70% by weight, free from catalyst, that had formed in other process steps, was fed through line (2). The liquid on the column bottom contained about 0.03% by weight of suspended solids predominantly consisting of terephthalic acid, which is preferably recovered.

A portion of the bottom liquid, containing 94% by weight of acetic acid, about 80% of catalyst fed to the column and oxidation intermediates, were recycled to the synthesis feed tank through pipe (3). A second portion of the bottom liquid passed to the column reboiler, subdivided into a still pot, equipped with a stirrer, and mto a heat exchanger, arranged in series and steam-heated. A liquid and concentrated synthesis purge was passed, through pipe (4), to a thin layer evaporator (not shown in the Figure) and then to an incinerator.

The remaining portion of the liquid leaving the pot passed through the exchanger and flowed back to the still pot, where the vapours were released and flowed back to the column bottom. The residence time in the lower part of the column, about 6 minutes, was not so long as to alter the organic substances of the recycled solution (3). The vapours flowing out from the column head, passed to a second (azeotropic) column for the acetic acid recovery, equipped with a reboiler, a reflux condenser and a demixing tank; into which an amount of azeotroping agent (isobutyl acetate) sufficient to make up for the losses, was introduced through pipe (5). A baffle easily separated the organic phase, which was returned to the column, from the aqueous phase, which collected in a trap and passed, through line (6), to the treatments for recovering the azeotroping agent. The acetic acid flowing out from the bottom contained 3.5% by weight of H2O and was passed, through line (7), to the synthesis feed tank, where it was mixed with the recylced liquid coming from line (3) and with fresh paraxylene coming from line (8).

32.3 parts by weight of electrolytic manganese were fed to an autoclave equipped with a heating jacket through pipe (10) from a proportioning hopper. 150 parts by weight of acetic acid were fed through pipe (9) thus converting (at 65"C and in about 8 hour petal manganese to the corresponding acetate. 231 parts of H20 containing 15.2 parts b lkeight of acetaldehyde were added through line (12). Finally 46 parts by weight of liqui romine at 99.8% were fed over a period of 45 minutes and the bromine was thoroughly d lkerted to hydrobromic ion without oxidation of manganese and precipitation of MnO2. At the conclusion of the elemental bromine reaction, tetrahydrated cobalt acetate was added in a Mn/Co weight ratio of 3/1. The resulting catalytic solution, having the following ratio: Br-/(Mn++ + Co++) = 1.09, entered, through line (13), the feed tank of terephthalic acid synthesis reactors.

Example 39 Utilizing the apparatus of Figure 2, a solution of manganese acetate and of cobalt acetate, in which the Mn:Co ratio was 3:1 by weight and in which bromine was absent, was fed, through line (13), to a storage tank for the feed of terephthalic acid to the synthesis. Tnto this tank, also a portion of the synthesis mother liquor, partially dried in the proper column and containing about 50% of the catalytic system contained in the original mother liquor, was introduced through line (3). Elemental bromine was added, through line (17), to the portion of partially anhydrified mother liquor flowing in line (3) and containing aldehydes (in particular carboxy-benzaldehyde) in amounts sufficient to prevent MnO2 from precipitating within the underlying tank.

The Br-:(Mn++ + Co++) ratio calculated in respect of cations Mn++ and Co++ coming from line (13), was 1.09 by weight, and the temperature of the liquid (3) was 90"C. The corresponding amount of bromine had to be increased by an amount sufficient to make up for the loss of bromine in the mother liquor (about 0.07 parts by weight of bromine for 1 part by weight of manganese and cobalt contained in the liquid recycled through line (3) ).

All the bromine was found quantitatively as hydrobrmic ion in synthesis feed line (14).

Claims

WHAT WE CLAIM IS:

1. A process for the synthesis of terephthalic acid, comprising oxidizing paraxylene in acetic acid solution in the presence of a catalytic system containing manganese, bromine and cobalt, in which bromine is added as elemental bromine, separating the synthesized solid terephthalic acid from the mother liquor and rectifying at least a portion of the mother liquor to reduce the water content therein in which: a) the catalytic system is prepared by bringing into contact acetic acid, manganese or manganese compounds, elemental bromine, a reducing compound, water and a soluble cobalt compound. b) the catalytic system obtained according to a) and containing manganese, bromine and cobalt in a dissolved form, is fed to the oxidation zone, optionally in admixture with the paraxylene feed and/or with at least a portion of the rectified mother liquor.

2. A process as claimed in Claim 1, in which the manganese is present in an amount of from 50 to 1(X)O parts per million with respect of acetic acid.

3. A process as claimed in Claim 1 or Claim 2 in which the manganese:cobalt weight ratio is from 2:1 to 4:1.

4. A process as claimed in any preceding claim in which the bromine:manganese weight ratio is a maximum 1.65.

5. A process as claimed in Claim 4 in which the bromine:manganese weight ratio is from 0.50 to 1.65.

6. A process as claimed in any preceding claim in which the acetic acid contains a maximum of 1% by weight of water.

7. A process as claimed in any preceding claim in which the reducing compound is acetaldehyde.

8. A process as claimed in Claim 7 in which the acetaldehyde:bromine molar ratio is at least 1.

9. A process as claimed in Claim 8 in which the acetaldehyde:bromine molar ratio is from 1 to 2.

10. A process as claimed in Claim 9 in which the acetaldehyde:bromine molar ratio is from 1.1 to 1.5.

11. A process as claimed in any one of claims 1 to 6 in which the reducing compound is diethyl-acetal.

12. A process as claimed in any one of claims 1 to 6 in which the reducing compound is ethylidene-diacetate.

13. A process as claimed in any one of claims 1 to 6 in which the reducing compound is the para-carboxybenzaldehyde , which is already present in the mother liquor of the paraxylene oxidation.

14. A process as claimed in any preceding claim in which the catalytic system is prepared at a temperature from 20 to 12() C.

15. , i process as claimed in Claim 14 in which the catalytic system is prepared at a tem ture from 40 to 90"C.

16. 5 process as claimed in any preceding claim in which the soluble cobalt compound is cobalt carbonate or cobalt acetate.

17. A process as claimed in any preceding claim in which the manganese is added in finely divided metallic form.

18. A process as claimed in any one of claims 1 to 10 in which the manganese is added in the form of manganese carbonate.

19. A process as claimed in any one of claims 1 to 16 in which the manganese is added as manganese acetate, optionally formed in situ from metallic manganese and acetic acid.

20. A process as claimed in any preceding claim in which at least a portion of the manganese and cobalt are in the form of a recovery mixture coming from the treatments of the purge of the terephthalic acid synthesis.

21. A process as claimed in preceding claim in which manganese, acetic acid and elemental bromine are initally brought into contact, in the presence of the reducing compound, and then the soluble cobalt compound is added.

22. A process as claimed in any of claims 1 to 20 in which acetic acid, metallic manganese and the reducing compound are initially brought into contact and then the other components are added.

23. A process for the synthesis of terephthalic acid substantially as herein described with reference to any one of Examples 9 to 39.

24. A process for the synthesis of terephthalic acid substantially as herein described with reference to the accompanying drawings.

25. Terephthalic acid prepared by a process as claimed in any preceding claim.