JP2900775B2 - Process for producing 2,6-naphthalenedicarboxylic acid - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION The present invention relates to a liquid phase oxidation of 2,6-diisopropylnaphthalene (hereinafter referred to as "2,6-DIPN") with molecular oxygen to give 2,6-naphthalenedicarboxylic acid. (Hereinafter referred to as “2,6-NDCA”). 2,6-ND
CA is used as a raw material for producing polyesters and polyamides, such as polyethylene naphthalate (PEN), which have excellent properties as engineering plastics, such as high heat resistance and high strength and low oxygen passage.

[0002]

2. Description of the Related Art Various methods for producing 2,6-NDCA by subjecting 2,6-DIPN to liquid phase oxidation with molecular oxygen in the presence of a catalyst comprising a heavy metal such as cobalt or manganese and a bromine compound have been proposed. And is publicly known. (Japanese Patent 60
-89445, JP-A-60-89446, JP-A-63-25
No. 0344, JP-A-1-121240, JP-A-1-1609
No. 43 gazette). JP-A-60-89445 discloses that 2,6-D
In order to oxidize IPN or its oxidation intermediate, the number of carbon atoms is 3
In a solvent containing at least 50% by weight of the following aliphatic monocarboxylic acid, in the presence of a heavy metal consisting of cobalt and / or manganese and a catalyst consisting of bromine, the heavy metal is 2,6-D
A liquid phase air oxidation process using at least 0.2 mole per mole of IPN or its oxidizing intermediate is proposed. It is further described that the proportion of 2,6-DIPN present in the oxidation reaction mixture should not exceed 0.2 mol per mol of heavy metal. JP-A-60-89446 proposes a method in which the heavy metal is present in an amount of at least 1% by weight based on the solvent. The point common to all of the above proposals, as typified by these, is that a heavy metal catalyst such as cobalt and manganese and a bromine compound are used in a large amount with respect to 2,6-DIPN. At the same time, all of the above proposals are oxidation reaction techniques by a semi-batch method. Semi-batch method means that a solution in which a catalyst is dissolved in advance is charged into a reactor,
A molecular oxygen-containing gas is continuously supplied, and 2,6-DIPN is slowly supplied thereto over a period of 1 to 4 hours. Thereafter, a post-oxidation reaction is further performed for 1 to 2 hours to complete the reaction. A reaction system in which a reaction product is taken out of the reactor after that.

On the other hand, 2,6-DIPN which is the object of the present invention
Are continuously supplied to the reactor while the reaction product is continuously withdrawn from the reactor, and are described in JP-A-62-212345 and JP-A-4-266846. Only. The method disclosed in JP-A-62-212345 relates to a process for continuously producing 2,6-NDCA, and comprises a continuous oxidation step-NDCA.
Crystal Separation Step-Catalyst Elution and Recovery Step in NDCA-N
It comprises a salting-out step of DCA and a step of recycling the recovered catalyst. Of these, in the continuous oxidation process, to suppress side reactions,
The amount of 2,6-DIPN or its oxidized intermediate during the reaction must be 0.4 or less, preferably 0.05 or less in molar ratio with respect to the total amount of cobalt and manganese heavy metal in the added soluble cobalt salt and manganese salt. It is stated that there is. Japanese Patent Application Laid-Open No. 4-266846 describes that the yield is improved by recycling a catalyst solution containing a reaction intermediate. In this case, it is described that it is appropriate to use a solution in which a catalyst containing Co, Mn, and Br is dissolved in an oxidizing raw material at least three times the weight. And the soluble cobalt salt and manganese salt for this reason, the total amount of heavy metals of cobalt and manganese is 0.03 to 100 g of the lower aliphatic monocarboxylic acid solvent in the generated solution.
It is added in an amount of 0.15 mol, preferably 0.04 to 0.12 mol. Further, it is described that 2,6-DIPN or its oxidation reaction intermediate is preferably supplied in an amount of 20 g or less per 100 g of the solvent. JP-A-4-266846 describes that by recycling a catalyst solution containing a reaction intermediate, the reaction intermediate yield can be improved as compared with the conventional method. In this, the catalyst concentration is suitably 0.02% by weight or more as a total of transition metals, and it is appropriate to use a solution in which a catalyst containing Co, Mn and Br is dissolved in an oxidizing raw material at least 3 times by weight. It has been described.

[0004]

In the conventional semi-batch production, the productivity is poor, the fixed cost is high, the production is not very industrial, and continuous production is indispensable.

[0005] In the case of a continuous production system, the reaction conditions are not constant, and it is difficult to stably perform a continuous reaction for a long time, and many by-products are generated. Further, a large amount of the catalyst is required, which causes many problems such as precipitation of the catalyst and clogging of internal parts of the reactor, deactivation of the catalyst itself, and loss of loss.

An object of the present invention is to solve the above problems and to provide a method for continuously producing 2,6-NDCA in high yield.

[0007]

Means for Solving the Problems In the conventional continuous production method, the control of the composition ratio of 2,6-DIPN and its oxidizing intermediate in the reaction solution, and the catalyst ratio for each individual substance, were not performed at all. The present inventors have found that by strictly controlling these composition ratios and the like, the amount of catalyst used can be significantly reduced, and a continuous reaction that is stable for a long time and extremely efficient can be realized. The present invention is based on these findings. That is, 2,6-DIPN
Is dissolved in a catalyst solution containing an aliphatic monocarboxylic acid in which a catalyst comprising cobalt, manganese and bromine is dissolved,
When continuously oxidizing with a gas containing molecular oxygen to produce 2,6-naphthalenedicarboxylic acid, the ratio of cobalt to manganese in the catalyst is set to be 0.1 ≦ Co / (Co + M) in atomic ratio.
n) ≦ 0.9, the ratio of bromine is 0.1 ≦ Br / (Co + Mn) ≦ 1 in atomic ratio with respect to the total of cobalt and manganese, and the concentration of 2,6-diisopropylnaphthalene in the reaction solution is the oxidation reaction intermediate. 0.1 in molar ratio to the concentration.
01 and a catalyst concentration of at least 20 times the molar ratio of 2,6-diisopropylnaphthalene to the total amount of the cobalt and manganese components of the catalyst and at least 10 times the molar ratio to the intermediate of the oxidation reaction. Thus, a continuous production method of 2,6-naphthalenedicarboxylic acid characterized by the following was able to stably perform a high yield reaction for a long time.

Further, the reaction product is continuously withdrawn from the reactor, the solid 2,6-NDCA produced is filtered off, and the catalyst solution containing the oxidation reaction intermediate is recirculated to the reactor to further efficiently improve the efficiency of the reaction. It was confirmed that 6-NDCA could be produced.

The 2,6-DIPN used as a raw material may contain a small amount of isomers such as 2,7-DIPN and other reaction by-products according to the production method and the like.

As an oxidation catalyst, a heavy metal comprising cobalt and manganese and a bromine compound are used in combination. Examples of cobalt and manganese include oxides, hydroxides, carbonates, organic acid salts, and halides. Of these, organic acid salts, particularly acetate salts, are preferred. Examples of the bromine compound include inorganic bromine compounds such as bromine (Br ₂ ), hydrogen bromide and hydrogen bromide, and alkyl bromides such as methyl bromide, ethyl bromide, propyl bromide and ethylene bromide. Among them, hydrogen bromide, sodium bromide, potassium bromide, ammonium bromide, cobalt bromide and manganese bromide are particularly preferred.

By using cobalt and manganese in combination, a higher catalytic action is observed as compared with the case where cobalt or manganese is used alone. The ratio of each heavy metal used as a catalyst is 0.1 ≦ Co / (Co + Mn) ≦ 0.
9, particularly preferably 0.2 ≦ Co / (Co + Mn) ≦ 0.7. Outside this range, the NDCA yield decreases.

The amount of the bromine compound, which is an oxidation reaction accelerator, relative to the total amount of cobalt and manganese metal is represented by an atomic ratio of 0.1 ≦ Br / (Co + Mn) ≦ 1, especially 0.2 ≦ Br / ( (Co + Mn) ≦ 0.5 is preferred. Below this range, the oxidation reaction does not proceed smoothly and the NDCA yield decreases. Above this range, an addition reaction of bromine to the naphthalene ring occurs, resulting in a decrease in 2,6-NDCA yield, a decrease in purity and quality deterioration, and an increase in the amount of methyl bromide entrained in the exhaust gas. Therefore, it is not preferable because a problem of air pollution occurs.

In the present invention, an aliphatic monocarboxylic acid is used as a solvent. Preferably, it has 3 or less carbon atoms, such as acetic acid and propionic acid. Mixtures with other solvents can be used. When a mixture of an aliphatic monocarboxylic acid and another solvent is used as the oxidizing solvent, the proportion of the aliphatic monocarboxylic acid in the solvent is preferably kept at least 50% by weight or more. If it is less than 50% by weight, the oxidation rate is suppressed, and the yield of NDCA is undesirably reduced. As the other solvent, a solvent stable to oxidation, such as benzene, monochlorobenzene, and monobromobenzene, is used.

As the molecular oxygen-containing gas, air can be used as it is. Further, oxygen or air diluted with an inert gas can be used. The oxygen partial pressure is preferably 0.2 to ² kg / cm ² .

In the continuous reaction, the raw material 2,6-DI is added to the oxidation reaction tank.
The reaction is carried out by continuously or intermittently supplying the PN, the catalyst solution, and the molecular oxygen-containing gas, while continuously or intermittently extracting the reaction solution in which 2,6-NDCA is generated and accumulated from the oxidation reaction tank. . Thereby, the liquid level in the oxidation reaction tank is always kept substantially constant. The extracted reaction solution is preferably subjected to solid-liquid separation of precipitated 2,6-NDCA, and the liquid portion is preferably returned to the reaction tank again.

In the method of the present invention, the reaction solution
The concentration of 2,6-DIPN is not more than 0.01 of the concentration of the oxidation reaction intermediate in a molar ratio, the concentration of the catalyst is not less than 20 times the molar ratio of the total amount of cobalt and manganese to the amount of 2,6-DIPN, and the molar ratio to the oxidation reaction intermediate The characteristic is that it is controlled so that it becomes 10 times or more.

This oxidation reaction intermediate is 2,6-DIPN
Is a generic term for all intermediates that lead to 2,6-NDCA by oxidation of Specifically, at least one isopropyl group of 2,6-DIPN is a 2-hydroperoxy-2-propyl group, a 2-hydroxy-2-propyl group, an acetyl group,
A typical example is a compound in which a formyl group or one of isopropyl groups has been oxidized to a carboxyl group.
These intermediates can be quantified by, for example, an ODS column and an internal standard method by high-performance liquid chromatography (HPLC) using water / acetonitrile as a developing solvent.

The 2,6-DIPN concentration of the reaction solution is not more than 0.01 of the concentration of the oxidation intermediate in a molar ratio, preferably about 0.01 to 0.0001, particularly preferably about 0.001 to 0.0001. The catalyst concentration is at least 20 times, preferably about 20 to 10,000 times, particularly preferably about 1000 to 10,000 times, the molar ratio of the total amount of cobalt and manganese to 2,6-DIPN.
It is about twice. The molar ratio of cobalt and manganese to the oxidation reaction intermediate is 10 times or more in total, preferably about 10 to 1000 times, and particularly preferably about 10 to 100 times.

The 2,6-DIPN concentration and the catalyst concentration in the reaction solution are within the above ranges by measuring the concentrations of 2,6-DIPN, the catalyst and the oxidation intermediate in the reaction solution continuously or at an appropriate time. The supply rate of 2,6-DIPN and the catalyst may be increased or decreased, or the reaction temperature, the reaction pressure and the like may be adjusted as much as possible. If it is known in advance that the above conditions will be satisfied if the reaction is continued under specific conditions, 2,6-D
Needless to say, measurement of IPN or the like is unnecessary.

The reaction can be carried out at a temperature in the range of 150 to 200 ° C.,
The temperature is more preferably 170 to 180 ° C. Below this range, the oxidation of 2,6-DIPN to the oxidation reaction intermediate proceeds, but the oxidation of the oxidation reaction intermediate to 2,6-NDCA does not proceed smoothly, and the yield of 2,6-NDCA decreases. I do. A higher reaction temperature is preferable because the reaction shifts to the product side (NDCA) and the concentration of unreacted 2,6-DIPN and its oxidation intermediate decreases, resulting in an increase in the concentration of the heavy metal catalyst relative to the oxidation intermediate. . However, when the reaction temperature is too high, the reaction pressure becomes too high, which increases the cost of the apparatus and further increases the combustion loss of the solvent.

The reaction pressure must be higher than the pressure at which the liquid phase can be maintained in the reactor, and an oxygen partial pressure of about 0.1 to 10 kg / cm ² is necessary in order to sufficiently provide the amount of oxygen required for the oxidation reaction. is necessary. Also, in order to sufficiently supply oxygen to the reactor to the liquid phase side, oxygen-containing gas is supplied through a nozzle, ring sparger, sintered metal, fine bubble diffuser made of porous porcelain, etc., to increase gas-liquid contact efficiency Is preferred.

The water generated by the oxidation reaction is gradually accumulated in the solvent and is allowed to some extent, but if it is present in a large amount, the catalytic activity decreases and the oxidation is inhibited. Therefore, it is necessary to remove generated water in the method of the present invention. Examples of this method include a method of separating water by condensing a gaseous phase gas and a method of separating water from a part of the circulating catalyst liquid by a distillation operation. That is, the present inventors first supply a solvent azeotropic with water, such as benzene and chlorobenzene,
A method was found in which the water generated by the reaction was entrained with the off-gas and extracted. In this method, a gas phase component entrained in the off-gas from the reactor is guided to a condenser (first condenser), a partially condensed liquid is circulated to the oxidation tank, and a non-condensed component is recycled to a subsequent condenser (second condenser). (Condenser) to completely condense and continuously withdraw from the oxidation tank out of the system to easily remove most of the oxidized water outside the system. It was also found that by newly replenishing an amount corresponding to a solvent such as acetic acid and an azeotropic solvent such as chlorobenzene which are removed at the same time, water generated by oxidation can be repeatedly used without accumulating in the catalyst solution. However, even if such an operation is not performed, it is also possible to use a general method in which a part of the circulating catalyst liquid is separated and removed by a distillation operation, and the method is not limited in the present invention. .

[0023]

The total amount of the cobalt and manganese catalysts is 2,6-
The reason why the control is such that it is present in a proportion of at least 20 times the moles of DIPN and at least 10 times the moles of the intermediate of the oxidation reaction is very early in the oxidation reaction, that is, from the isopropyl group to the hydrolyzate. This is because a very large amount (at least 20 times the amount of the heavy metal catalyst required for 2,6-DIPN is required for the 2,6-DIPN) in the step of producing peroxides, hydroxy groups, and the like is required. In the step of oxidizing to the group, the reaction proceeds stably if a heavy metal catalyst is present at least 10 times the mole of the oxidation reaction intermediate. If the amount of the heavy metal catalyst is 20 times or less with respect to 2,6-DIPN or 10 times or less with respect to the oxidation reaction intermediate at the beginning of the reaction, the 2,6-NDCA yield decreases, and trimellitic acid Not only is the amount of by-products significantly increased, but trimellitic acid is not preferable because it forms a complex with heavy metals such as cobalt and manganese, and the amount of catalyst loss increases.

In order to reduce the concentration of the catalyst in the solution and the amount of the catalyst to be used, the concentration of 2,6-DIPN in the reaction solution should be as low as possible with respect to the intermediate of the oxidation reaction, preferably the intermediate of the oxidation reaction. It is desirable to keep the molar ratio to 0.01 mol times or less. The reason for keeping the catalyst concentration low is not only that it is inexpensive in terms of cost, but also because the possibility of clogging in a pipe or the like due to the precipitation of the catalyst is reduced, and the operation becomes easy.

The solvent containing the oxidation reaction intermediate and the catalyst obtained after removing the reaction product coming out of the reactor can be repeatedly used by circulation, whereby the oxidation reaction intermediate is finally converted to 2,2. Because it is oxidized to 6-NDCA,
The 2,6-NDCA yield increases as compared to the case where the catalyst solution is not circulated.

[0026]

The present invention will be specifically described below with reference to examples and comparative examples. The following examples were carried out using a titanium-lined 1.5 l oxidation reactor shown in FIG.

Example 1 First, 0.9 l of acetic acid was previously charged in the oxidation reaction tank 1,
Air pressurized by a compressor was supplied from the air supply line 9 at 4 Nl / min. The inside of the oxidation reaction tank 1 was stirred at 1000 rpm by a disk turbine type stirring blade, heated from the outside by a heating furnace and maintained at 170 ° C., and at the same time, adjusted by the pressure holding valve 6 to adjust the inside of the reaction tank to 8 kg / cm ² (gauge pressure). ).

249.0 g of cobalt acetate (tetrahydrate), 245.0 g of manganese acetate (tetrahydrate), 51.5 g of sodium bromide and 2850.0 g of acetic acid were charged into the catalyst circulation tank 10 and heated to 80 ° C. . Next, 300 g / hr of the catalyst solution was supplied to the oxidation reaction tank 1 via the catalyst circulation line 11 by the high-pressure pump 17.
Supplied. At the same time, the extraction of the product from the product extraction line 12 was started. The product is withdrawn by a withdrawal valve according to an instruction from a liquid level controller provided in the oxidation reaction tank 1.
Automatic control of opening and closing of 18 was performed, and the liquid level in the oxidation reaction tank was discharged so as to keep it constant (internal liquid volume 0.9 l).

The outlet gas temperature of the first condenser 2 was controlled by flowing the cooling water so that the temperature was controlled to 10 ° C., so that all the gas phase components accompanying the off-gas were condensed. Components not condensed in the first condenser were further cooled in the second condenser 3 and completely condensed and trapped.

2,6-DIPN was dissolved by heating at 80 ° C.
The reaction was supplied to the oxidation reaction tank 1 at 0 g / hr to start the reaction. The catalyst solution containing solid NDCA discharged from the withdrawal line 12 after the start of the reaction was discarded as it was until 10 hours after the start of the reaction. At the time when 10 hours have elapsed, half the amount of cobalt acetate tetrahydrate and manganese acetate tetrahydrate charged at the start of the reaction,
Sodium bromide and acetic acid (124.5 g, 122.5 g, 25.8 g, respectively)
g, 1425.0 g) was newly added to the catalyst circulation tank 10. From the eleventh hour, the reaction product discharged from the product withdrawal line 12 after the start of the reaction was separated into solids (crude NDCA) by a filter 13 and 1 kg of a filtrate (a catalyst solution containing an oxidation reaction intermediate) was collected. At this time, 260 g of the solvent was evaporated using an evaporator to remove an amount of water corresponding to 10.2 g of theoretically generated water. Thereafter, 250 g of acetic acid was additionally replenished, and returned to the catalyst circulation tank 10 again for repeated use. As a result, more water than the theoretical amount generated by the oxidation reaction could be removed from the system, and as a result, the water content in the circulating catalyst solution could be kept at 3 to 5%.

From about 20 hours after the start of the reaction, the amount of the reaction product, the composition of the reaction solution, the water concentration in the reaction solution, and the like were all stabilized.

By performing the above operation, a stable reaction was able to be performed from the 20th hour to the 100th hour after the start of the continuous operation by the catalyst circulation until the reaction was completed.
During this time, when the catalyst liquid discharged from the catalyst circulation line 12 (equivalent to the liquid composition during the reaction because it is in a completely mixed state) was taken out and analyzed, it was found that only about 0.1 mmol / l of unreacted DIPN was present. The molar ratio to the reaction intermediate is
It was 0.002-0.003. It was also found that the total amount of cobalt and manganese relative to the oxidation reaction intermediate was 15 to 16 times in molar ratio. The yield of NDCA (based on DIPN) was 82% at first, but the yield began to improve when the catalyst circulation was continuously continued, and was stabilized at about 90 to 93% for 20 to 100 hours.
No decrease in the activity of the circulating catalyst was observed. Crude NDCA
Was a light yellow powder that was smooth and had good solid-liquid separation properties from the catalyst. The trimellitic acid yield was 6%.

Comparative Example 1 83.0 g of cobalt acetate (tetrahydrate) was placed in a catalyst circulation tank 10.
Manganese acetate (tetrahydrate) 81.7 g, sodium bromide 17.2
After charging g and 3110.0 g of acetic acid, the experiment was started in exactly the same manner as in Example 1. After 10 hours, half of the cobalt acetate tetrahydrate, manganese acetate tetrahydrate, sodium bromide, and acetic acid charged at the start of the reaction were newly added to the catalyst circulation tank 10. Thereafter, continuous operation was performed by circulating the catalyst in the same manner as in Example 1. The system became stable from the time when 20 hours had passed since the start of the reaction.

After the system was stabilized, the catalyst liquid discharged from the catalyst circulation line 12 was taken out and analyzed.
IPN is present at about 0.1 mmol / l as in Example 1, and the molar ratio to the oxidation reaction intermediate is 0.007 to 0.010.
Met. On the other hand, it was found that the total of cobalt and manganese relative to the oxidation reaction intermediate was present 7 to 8 times in molar ratio. The reaction was terminated 50 hours after the start of the reaction.
The NDCA yield up to 20-50 hours after the start of the reaction is 35-40%
And it was quite low. The crude NDCA was a greasy dark yellow powder, which was clearly different in properties from that obtained in Example 1. When the reactor and piping were inspected after the completion of the reaction, considerable adhesion of tar-like products was observed.

Example 2 The inside of the oxidation reaction tank 1 was maintained at 200 ° C. and 30 kg / cm ² (gauge pressure), and cobalt acetate (tetrahydrate) 83.0
g, manganese acetate (tetrahydrate) 81.7 g, sodium bromide
17.2 g and 3110.0 g of acetic acid were charged, and the experiment was started in exactly the same manner as in Example 1. 2,6-DIPN at 80 ° C
And heated to 30.0 g / hr and supplied to the oxidation reaction tank 1 to start the reaction. The catalyst solution containing solid NDCA discharged from the withdrawal line 12 after the start of the reaction was discarded as it was until 10 hours after the start of the reaction. At the end of 10 hours, half the amount of cobalt acetate tetrahydrate, manganese acetate tetrahydrate, sodium bromide and acetic acid (41.5
g, 40.9 g, 8.6 g, and 1555.0 g) were newly added to the catalyst circulation tank 10. From the 11th hour, after the reaction is started, the reaction product discharged from the product withdrawal line 12 is separated into solids (crude NDCA) by a filter 13 and then 1 kg of a filtrate (a catalyst solution containing an oxidation reaction intermediate) is accumulated. At this point, 300 g of the solvent was evaporated using an evaporator, and the amount of water corresponding to 15.3 g of theoretically generated water was removed. Thereafter, 285 g of acetic acid was additionally replenished, and returned to the catalyst circulation tank 10 again for repeated use. As a result, more water than the theoretical amount generated by the oxidation reaction can be discharged out of the system, and as a result, the water contained in the circulating catalyst solution is 4 to 6
%.

By performing the above operation, a stable reaction was able to be performed from the 20th to the 100th hour after the start of the continuous operation by the catalyst circulation until the reaction was completed.
During this time, the catalyst liquid discharged from the catalyst circulation line 12 was taken out and analyzed, and it was found that unreacted DIPN was present in less than 0.01 mmol / l. The total amount of cobalt and manganese with respect to the oxidation reaction intermediate dissolved in the liquid during the reaction was about 12 times in molar ratio. NDCA yield is about 93 ~
The activity of the circulating catalyst was not reduced during the reaction time. Further, the crude NDCA obtained in this reaction was a smooth yellow powder like the product obtained in Example 1, and the solid-liquid separation property with the catalyst was good. The trimellitic acid yield was 4%.

Comparative Example 2 In a catalyst circulation tank 10, 33.2 g of cobalt acetate (tetrahydrate) was added.
Manganese acetate (tetrahydrate) 32.7 g, sodium bromide 7.0 g
The experiment was started by the same operation as in Example 2 except that 3220.0 g of acetic acid and 3220.0 g of acetic acid were charged. After 10 hours, half of the cobalt acetate tetrahydrate, manganese acetate tetrahydrate, sodium bromide, and acetic acid charged at the start of the reaction were newly added to the catalyst circulation tank 10. Thereafter, continuous operation was performed by circulating the catalyst in the same manner as in Example 2.

After the system was stabilized, the catalyst liquid discharged from the catalyst circulation line 12 was taken out and analyzed.
Only about 0.2 mmol / l of IPN was present, and the molar ratio to the oxidation reaction intermediate was 0.007 to 0.010. On the other hand, it was found that the total of cobalt and manganese relative to the oxidation reaction intermediate was present in a molar ratio of 2 to 3 times.

At 38 hours after the start of the reaction, the reaction was stopped because the liquid level in the reactor could not be controlled due to the blockage of the extraction valve 18. The NDCA yield from 20 to 38 hours after the start of the reaction was 38% on average, and it was a greasy dark yellow powder, which was clearly different from those obtained in Example 2.

Example 3 The inside of the oxidation reaction tank 1 was maintained at 150 ° C. and 8 kg / cm ² (gauge pressure), and cobalt acetate (tetrahydrate) 83.0
g, manganese acetate (tetrahydrate) 81.7 g, sodium bromide
17.2 g and 3110.0 g of propionic acid were charged, and the experiment was started in exactly the same manner as in Example 1. 2,6-DIP
N was heated and dissolved at 80 ° C. and supplied to the oxidation reaction tank 1 at 5.0 g / hr to start the reaction. The catalyst liquid containing solid NDCA discharged from the withdrawal line 12 after the start of the reaction
Until the hour, it was discarded. At the end of 10 hours, half of the cobalt acetate tetrahydrate, manganese acetate tetrahydrate, sodium bromide and propionic acid charged at the beginning of the reaction (41.5 g, 40.9 g, 8.6 g, 1555.0 g respectively) (Equivalent) was newly added to the catalyst circulation tank 10. After the reaction starts from the 11th hour, the reaction product discharged from the product withdrawal line 12 after the solid (crude NDCA) is separated by the filter 13 and when 1 kg of the catalyst solution containing the oxidation reaction intermediate has accumulated is obtained. Using an evaporator, 286.5 g of the solvent was evaporated to remove an amount of water corresponding to 2.6 g of theoretically generated water. Thereafter, 280 g of propionic acid was additionally replenished, and returned to the catalyst circulation tank 10 again for repeated use. As a result, more water than the theoretical amount generated by the oxidation reaction could be discharged out of the system, and as a result, the water content in the circulating catalyst solution could be kept at 1% or less.

By performing the above-mentioned operation, a stable reaction was able to be performed from the 20th hour to the 100th hour after the start of the continuous operation by circulating the catalyst until the reaction was completed.
During this time, the catalyst liquid discharged from the catalyst circulation line 12 was taken out and analyzed. As a result, only about 0.3 mmol / l of unreacted DIPN was present, and the molar ratio to the oxidation reaction intermediate was about 0.01. It was also found that the total amount of cobalt and manganese relative to the oxidation reaction intermediate was 15 to 16 times in molar ratio. The NDCA yield was 83-85%, and no decrease in the activity of the circulating catalyst was observed during the reaction time. The crude NDCA obtained by this reaction was somewhat colored compared to the product obtained in Example 1, but was a smooth yellow powder and had good solid-liquid separation properties from the catalyst. The trimellitic acid yield was 13%.

Comparative Example 3 83.0 g of cobalt acetate (tetrahydrate) was added to the catalyst circulation tank 10.
Manganese acetate (tetrahydrate) 81.7 g, sodium bromide 17.2
Example 3 except that g and 3110.0 g of propionic acid were charged.
The experiment was started by the same operation as described above. After 10 hours, half of the cobalt acetate tetrahydrate, manganese acetate tetrahydrate, tonalium bromide, and propionic acid, which were charged at the beginning of the reaction, were newly added to the catalyst circulation tank 10.
Thereafter, continuous operation by circulating the catalyst by the same operation as in Example 3 was performed.

After the system was stabilized, the catalyst liquid discharged from the catalyst circulation line 12 was taken out and analyzed.
IPN was present at less than 0.01 mmol / l and the molar ratio to the oxidation reaction intermediate was 0.005-0.01. On the other hand, it was found that the total amount of cobalt and manganese relative to the oxidation reaction intermediate was 5 to 6 times in molar ratio.

After 40 hours from the start of the reaction, the reaction was stopped because the liquid level in the reactor could not be controlled due to the blockage of the withdrawal valve 18. The average yield of NDCA for 20 to 40 hours after the start of the reaction was about 35%, and it was a sticky dark yellow powder, which was clearly different in properties from that obtained in Example 3.

[0045]

According to the present invention, the total amount of the cobalt and manganese catalysts in the reaction solution is 20 to 6-DIPN.
By performing the continuous oxidation reaction while controlling it to be present at a ratio of at least 10 times the molar amount of the intermediate of the oxidation reaction and at least 10 times the molar ratio of the intermediate of the intermediate, it is possible to obtain 2, 6-NDCA could be obtained.

The 2,6-DIPN present in the reaction solution
By controlling the reaction so that the concentration was as low as possible with respect to the concentration of the oxidation intermediate, the catalyst concentration could be further reduced.

Further, the oxidation reaction is continuously carried out according to the above method, solid NDCA is filtered off from the product continuously withdrawn from the reactor, and the catalyst solution containing the oxidation reaction intermediate obtained thereafter is returned to the reactor. 2,6-NDC can be obtained in a higher yield by recycling than in the case of not recycling.
A could be produced. When the catalyst solution containing the oxidation reaction intermediate is recycled to the reactor, the oxidation reaction intermediate is repeatedly returned to the reactor and finally oxidized to 2,6-NDCA or insoluble by-products. There is no going out of the system. Therefore, the 2,6-NDCA yield is improved as compared with the case where the catalyst solution containing the oxidation reaction intermediate is not recycled.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a reaction apparatus used in an example of the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 ... Oxidation reaction tank 2 ... 1st condenser 3 ... 2nd condenser 4 ... 1st tank 5 ... 2nd tank 6 ... Holding pressure valve 7 ... Off gas discharge line 8 ... DIPN supply line 9 ... Air supply line 10 ... Catalyst circulation Tank 11 ... Catalyst circulation line 12 ... Product extraction line 13 ... Filter 14,15 ... Valve 16,17 ... High pressure pump 18 ... Extraction valve

──────────────────────────────────────────────────続 き Continuation of front page (72) Inventor Mitsuki Yasuhara 1-2-1, Marunouchi, Chiyoda-ku, Tokyo Inside Nippon Kokan Co., Ltd. (56) References JP-A-62-212345 (JP, A) (58) Survey ⁶ (Int.Cl. ⁶ , DB name) C07C 63/38 B01J 31/30 C07C 51/265 C07B 61/00 300

Claims (1)

(57) [Claims] (1) 2,6-diisopropylnaphthalene,
When continuously producing 2,6-naphthalenedicarboxylic acid by dissolving in a catalyst solution containing an aliphatic monocarboxylic acid in which a catalyst consisting of cobalt, manganese and bromine is dissolved and oxidizing with a gas containing molecular oxygen, The ratio of cobalt and manganese in the catalyst is 0.1 ≦ C in atomic ratio.
o / (Co + Mn) ≦ 0.9, and the ratio of bromine is 0.1 ≦ Br / (Co
+ Mn) ≦ 1, and the concentration of 2,6-diisopropylnaphthalene in the reaction solution is 0.01 in molar ratio to the concentration of the oxidation reaction intermediate.
And the catalyst concentration is at least 20 times the molar ratio of 2,6-diisopropylnaphthalene to the total amount of the cobalt and manganese components of the catalyst, and 10 times the molar ratio to the oxidation reaction intermediate.
Method for continuous production of 2,6-naphthalenedicarboxylic acid, characterized in that the amount is twice or more