NO150679B - PROCEDURE FOR THE PREPARATION OF TERPHALIC ACID BY THE OXIDATION OF AN EXTRAORDINABLE OUTPUT MATERIAL FOR TERPHALIC ACID IN THE PRESENT OF MANGANIC AND / OR COBAL SALTS - Google Patents

Description

The present invention relates to a method

for the production of terephthalic acid by oxidation of a reaction mixture, which contains at least one oxidizable starting material for terephthalic acid in the form of p-toluene acid, mixtures of p-toluene-

acid and p-xylene and/or partially oxidized p-xylene derivatives and 5-80% by weight of water in the presence of an oxidation catalyst containing at least one catalytically active metal compound in the form of manganese or cobalt salts of carboxylic acid and optionally an inert organic solvent in an amount of 0-100% of water quantity

it, at a temperature of between 140 and 220°C and a pressure which is sufficient to keep at least part of the water in the liquid phase and which lies at o between 5 and 40 kg/cm 2.

Terephthalic acid is of great commercial importance as it i

increasingly used as starting material for the production of resins with a high molecular weight, e.g. fiber and film-forming polyesters.

Many methods are known for liquid phase oxidation

of alkyl-substituted aromatic compounds to aromatic carboxylic acids. One of the first patents in this area is US-

patent 2,245,528, which describes a one-step process for the oxidation of alkylaromatic compounds using mole-

chiral oxygen in the presence of a metal catalyst, a solution

agent, e.g. acetic acid, and possibly an oxidation initiator.

But even under strict conditions, the yield of dicarbon-

low acidity. E.g. by oxidation of a mixture of xylenes and air in acetic acid, containing cobalt and manganese acetates as catalysts, at 185-200°C under a pressure of 50 atmospheres and in the presence of diethyl ketone as initiator, the yield of phthalic acids was only 2%, and the main reaction products were toluene acids and other oxidation intermediates.

Methods for oxidizing p-xylene in one step with a higher yield of terephthalic acid are known from several other patents. These patents mainly indicate the use of special activators, e.g. bromine-containing compounds (US Patent 2,833,816), ketones (US Patent 2,853,514) or aldehydes (US Patent 3,036,122). It is true that certain of these processes are practiced commercially, but in spite of this they are fraught with serious disadvantages. Thus, serious corrosion problems arise when using bromine-containing activators. When a ketone or aldehyde is used, a part of these is inevitably lost, and the remaining part is mainly converted into acetic acid, which must be separated, purified and used commercially for the process to be economically viable. However, despite these disadvantages, the use of an activator is considered to be an essential prerequisite for an efficient production of phthalic acids from xylenes.

In most cases, it is also stated that it is necessary to use a solvent. Fatty acids with a low molecular weight, specifically acetic acid, are used to a large extent for this purpose. The added amount of solvent must be sufficient to keep the reacting substances and reaction products in solution or at least in suspension without difficulty. In this way, the stirring of the reaction mixture is facilitated, the distribution of oxygen is improved, the formation of by-products is reduced to a minimum and the heat of reaction is easily removed by evaporation of the solvent. However, under the reaction conditions that are usually used, a significant amount of solvent is lost by co-oxidation. In addition, the solvent must be separated from the other components of the reaction mixture and then cleaned and recycled. This consumption of part of the solvent and these operations for the recovery of the remaining part are associated with additional treatment costs.

In order to avoid the above-mentioned significant problems, it is proposed to carry out the oxidation of p-xylene without a solvent. In this case, it is of course necessary to work at a temperature that is at least in the range of p-toluene acid's melting point, i.e. approx. 180°C, in order to obtain a liquid reaction mixture. The choice of temperature is therefore limited. It is also true that without a solvent the handling of the reaction mixture and the separation and purification of the terephthalic acid are difficult to carry out. Usually the reaction is carried out to the point where the terephthalic acid content of the mixture does not exceed 60, preferably 45% by weight. Beyond this point, "it becomes difficult to handle the oxidation reaction mixture as a slurry, so that the reaction operation is adversely affected" (quotation, see US Patent 3,883,584).

Without a solvent, the removal of the heat of reaction entails another difficulty in manufacturing on an industrial scale, as extensive fouling takes place in the reaction vessel, even if terephthalic acid does not occur in large quantities. Thus, it is stated in US patent 2,696,499, which relates to the oxidation of xylene to toluene acids in the absence of a solvent, that "the most important design problem in connection with the cooling of the xylene oxidation mixture is to prevent solids from settling on the container walls".

From US patent 3,406,196, a two-stage process is known in which water is used as a suspending agent for terephthalic acid. In the first step, an alkyl aromatic compound, more specifically p-xylene, is oxidized with air in the absence of a special solvent to partially oxidized compounds, which in the second step are further oxidized at a higher temperature in the presence of significant amounts of water as suspension medium. Bromine or a bromine-containing compound must thereby be included to promote the oxidation. Nevertheless, very high temperatures, in the range of 200-275°C, preferably 225-250°C, are required to achieve conversion of these partially oxidized compounds to terephthalic acid. Consequently, one naturally encounters the same or even worse corrosion problems than those that occur in processes where acetic acid is used as a solvent. Moreover, as stated in the last-mentioned patent document, it is true that "there is a tendency for significant losses of unreacted polyalkylaromatic compounds to occur due to decomposition and other side reactions, when such compounds are exposed to the high temperatures found necessary for an efficient conversion of those in the first oxidation step formed partially oxidized compounds to aromatic polycarboxylic acids". It is clear that according to this patent the use of large amounts of water, even in the presence of a bromine-type accelerator, does not give satisfactory results for the oxidation of p-xylene to terephthalic acid in one step.

In fact, it has long been known that water is "a catalyst poison in oxidation reactions" (US Patent 2,696,499). According to the most common belief, water has an unfavorable effect on the reaction rate by interfering with the initiation of the reaction. As a general rule, its presence should be avoided as much as possible, regardless of whether a solvent and/or an activator is present or not. From US patent 3,064,044, a technical method is thus known for maintaining a final oxidation (bromine-accelerated) under approximately anhydrous conditions. In US patent 3,519,684, which relates to an oxidation process where peracetic acid is used as an accelerator, it is stated that approximately anhydrous conditions are preferably used, although a water content of up to 10% may be permitted and a maximum water content of no more than 5 % is preferred. In a continuous process for the oxidation of xylenes in the absence of any accelerator and where the partially oxidized intermediates are continuously recycled, water is separated from the waste liquid before it is recycled in order to "maintain the water content of the reaction mixture at less than 15%, preferably less than 5% of the total amount reaction mixture" (citing US Patent 3,700,731).

In recent times, it has quite unexpectedly been shown that the oxidation of p-xylene to terephthalic acid can be carried out in the presence of significant amounts of water as solvent, but without any bromine-containing activator which was previously considered essential and necessary to use. This process is described in Norwegian patent specification 147,600 and US patent specification 4,278,810. From these it is known oxidation of p-xylene liquid phase by means of an oxygen-containing gas in the presence of p-toluene acid, water and a heavy metal salt as catalyst at a temperature of between 140 and 220°C and under a pressure which is sufficient to keep at least part of the water in liquid phase. Since, however, the mutual solubility of p-xylene and water is low at the working temperature, such mixtures of water, p-xylene and p-toluene acid can be divided into two phases: a water phase and an organic phase which is rich in hydrocarbons and also contains a significant part of the p-toluene acid which is part of the reaction mixture. In this case, the oxidation reaction takes place mainly in the organic phase, where the concentration of water is relatively low. This phase separation also entails significant technical difficulties in terms of homogenisation, oxygen distribution and mass transfer effects.

The purpose of the present invention is to produce a method for the production of terephthalic acid, where this is achieved with a high yield and a high degree of purity and where the above-mentioned disadvantages of known methods are avoided.

According to the invention, this is achieved by using water in such an amount that the liquid reaction mixture is practically homogeneous, and that the catalyst is used in an amount that is sufficient to give a minimum amount of M millimoles of the catalytically active metal compound per kg of the liquid reaction mixture, where M is defined by the following equation:

where y is the molar ratio between water and p-toluene acid in reaction M

the mixture, x is the molar ratio Mn, and

A = 0.2 00 + 0.031, <M>Mn <+><M>Co

B = 10.9 + 1.3,

C = 4.35 + 0.17,

D = 0.0724 + 0.0117.

According to the method, it is not necessary to use special corrosion-resistant equipment, as it can be carried out in an equipment made of ordinary stainless steel.

There is no need to use any accelerator, e.g. bromine compound, in addition to the oxidation catalyst.

According to the method, the terephthalic acid can be easily recovered from the oxidized reaction mixture at relatively moderate temperatures.

The catalyst and oxidation intermediates can be recovered and reused for oxidation.

According to the invention, the terephthalic acid can be produced on an industrial scale at relatively low costs.

The procedure can be carried out both batchwise and continuously.

The quantities of catalyst, water and p-toluene acid which are necessary to ensure an effective oxidation of the starting material can be easily calculated.

By keeping the catalyst concentration at at least the minimum value specified above, an effective oxidation of any of the starting material mixtures specified above to terephthalic acid is achieved in a homogeneous reaction mixture in the liquid phase, whereby it is not necessary to use solvents other than water.

The invention will be further explained in the following detailed description of preferred embodiments with reference to the accompanying drawing showing a phase diagram for mixtures of p-xylene, p-toluene acid and water at a temperature of 185°C.

An important aspect of the present invention is that the harmful effect of water, to which, according to prior art, attention has been paid, can be avoided if the oxidation is carried out under the above specified special conditions. It can be carried out for the oxidation of various materials which can consist exclusively of p-toluene acid or in a mixture with p-xylene and/or partially oxidized derivatives thereof, e.g. p-tolualdehyde. Moreover, these materials can be oxidized in a homogeneous water solution if the amount of catalyst is chosen in relation to the ratio between the amounts of water and p-toluene acid included in the system.

The method according to the invention can be carried out in batches or continuously. The substances that participate in the reaction are dissolved in water, and the catalyst is added to the resulting solution. Oxygen is then introduced into the mixture, and the oxidation is carried out in the above-mentioned temperature range of between 140 and 220°C under the indicated pressure. The pressure is regulated in such a way that the reaction mixture at the working temperature is kept at least approximately in the liquid phase. Terephthalic acid separates from the reaction mixture in the form of a white, crystalline precipitate. In a continuous process, this precipitate is continuously removed by a suitable method for separating solids from liquid, e.g. filtration, centrifugal separation or sedimentation and decantation, and the remaining liquid, which contains unreacted reaction products and oxidation intermediates, is recycled to the oxidation area. New reaction participants are continuously added as compensation for the terephthalic acid thus separated and for by-products formed. Thus, there is no need to add chemicals from the outside. Nevertheless, when carrying out the method according to the invention, one or more organic solvents can be added to the reaction mixture which do not interfere with the reaction and which are relatively inert under the working conditions. As examples of such compounds that can be used as a solvent in a mixture with water, benzoic acid and acetic acid can be mentioned. These substances, which are actually formed in smaller quantities during the reaction, can thus be allowed to accumulate to a certain, smaller extent in the reaction mixture. The advantage achieved by the present invention is, however, achieved regardless of the presence of such substances, which, if used, should not be included in an amount that exceeds the amount of water in the system.

The amount of water that is appropriately used in the method according to the invention can vary within wide limits,

as indicated above between 5 and 80% of the weight of the reaction mixture, depending on various factors. As already stated, an essential feature of the invention is that the oxidation is carried out in a homogeneous, water-based system. For this reason, the water quantity is chosen

it so that it is sufficient to give an approximately homogeneous water solution of the substances that are to react with each other, taking into account the other process variables, such as temperature

turn, and the relative amounts of the various compounds to be oxidized. When the method e.g. must be used for the oxidation of p-toluene acid, exclusively or in a mixture with other oxygen-containing compounds which are relatively soluble in water, e.g. p-tolualdehyde, a quantity of water is selected which is sufficient to completely dissolve p-toluene acid at the working temperature. As the solubility of p-toluene acid in water increases strongly with temperature, the amount of water used can be reduced with increasing temperature. A general rule, however, is that the amount of water is not less than 5%, preferably 10%, of the weight of the reaction mixture.

When p-xylene is also included as a component of the reaction mixture, the amount of water should not exceed an amount where separation of the mixture into two liquid phases will take place. This amount is of course dependent on the amount of p-xylene that in-

goes into the mix.

The drawing shows a triangular phase diagram for mixtures

of p-xylene, p-toluene acid and water at a temperature of 185°C

(calculated in weight%). According to this diagram, the system in zone A is a homogeneous solution, while in zone B it is two-phase. A professional on

the area will easily find that the boundary between these two zones does not vary much with temperature. It can be seen that the amount of p-xylene must be limited to avoid the presence of a significant organic phase.

When for that reason the oxidation of p-xylene in the process

according to the invention is carried out in batches, p-xylene should be added successively, either intermittently or continuously, to the reaction mixture in such balanced quantities that the system is kept in zone A.

According to a preferred embodiment of the invention, the reaction is carried out in a continuous flow process where unreacted p-xylene and the oxidation intermediates, i.e. mainly p-toluene acid, are continuously recycled into the reaction zone. In this case, the molar ratio between p-toluene acid and p- of xylene in the reaction mixture in the equilibrium state will be in the range between 3 and 15, mainly depending on the temperature, or expressed differently: the reaction mixture will consist of the shaded area of the diagram in the drawing.It appears that this area lies almost entirely in zone A, i.e. corresponds mainly to homogeneous solutions. From what has been mentioned above, it will be clear that according to the invention it is always possible to regulate the temperature and/or the amount of water used as solvent in such a way that a homogeneous system is achieved. If the method according to the invention is carried out continuously, the continuous process known from US patent 4,278,810 is preferred.

But other factors must also be taken into account. Since terephthalic acid, the desired end product, is largely insoluble

in the reaction mixture, a sufficient amount of water must thus be added in order to achieve a workable slurry. However, there is no advantage in using such a large amount of water that more than e.g. 10% terephthalic acid dissolves at the working temperature. Another factor to consider is the rate of reaction: it

it is certainly within the scope of the invention to carry out the oxidation reaction in a medium containing up to 80% by weight of water, but the presence of an excessively large amount of water can have a negative effect on the reaction rate. Moreover, in some cases, a part of the catalyst can deviate from its catalytic function by forming a black compound that precipitates out of the reaction medium. Furthermore, for economic reasons, it is disadvantageous to lose part of the reaction vessel's capacity by using an unnecessarily large amount of water. In general, therefore, for these various reasons, the amount of water in the system should not exceed 75, preferably 60% of the weight of the reaction mixture.

The oxidation reaction is carried out at a temperature of at least 140°C. Below this temperature, it is difficult to use the reaction mixture in the form of a homogeneous solution. On the other hand, the reaction at a temperature of over 220°C will result in increased overoxidation, undesirable side reactions and corrosion problems. In most cases, a reaction

temperature of between 150 and 190°C.

The pressure is regulated as a function of the temperature. It is necessary to use a pressure which is sufficiently much higher than the atmospheric pressure in order for the reaction mixture to be maintained in liquid form at the working temperature. A pressure that exceeds this value is usually applicable to ensure active oxidation. As mentioned, the pressure is kept between 5 and 40 kp/cm.

The oxidation catalyst used in the method according to the invention is at least one metal compound in the form of manganese and/or cobalt salts of carboxylic acid, provided that it is at least partially soluble in the aqueous reaction mixture or is capable of forming a soluble or at least soluble connection with one of the reaction participants in the mixture. Examples of such carboxylic acid salts are acetates, naphthenates, toluates and the like. In addition to cobalt and/or manganese compounds, the catalyst can contain nickel salts.

A fundamental feature of the present invention is that the minimum catalyst concentration that is necessary to ensure oxidation in a homogeneous, water-based system is dependent on the respective amounts of water and p-toluene acid included in the system. In a homogeneous water solution of p-toluene acid and possibly p-xylene and/or partially oxidized derivatives thereof, it is a fact that oxidation cannot take place if the amount of active catalyst is below a critical concentration M in millimoles of the metal compound per kilogram of the reaction mixture, given by the above equation:

where y is the molar ratio between water and p-toluene acid, x is the amount of manganese in the metal catalyst in parts per moles in relation to the total amount of manganese and cobalt in the catalyst, i.e. M Mn,

A = 0.200 + 0.031, B = 10.9 + 1.3, <M>Mn <+><M>Co

C = 4.35 + 0.17 and D = 0.0824 + 0.0117.

The value of M, as it is calculated according to equation (1) for a given value of y and x, must thus be considered to be an estimate of the real critical concentration, which, statistically expressed, lies in a confidence interval with M as the centre. As those skilled in the art will appreciate, this confidence interval is dependent on the values of y and x used in calculating M, but in each particular case it can be calculated using the above data using Taylor's formula. Thus, e.g. that the critical catalyst concentration will lie in the confidence interval 21.46 + 0.81 if y = 70 and x = 1.0 and in the interval 2.74 + 0.45 if y=logx=0.5.

The critical catalyst concentration defined above

is the lowest that can be used under given conditions. At lower concentrations, the process cannot be carried out. In practice, it is of course preferred to use a catalyst amount that is above this minimum value. The oxidation rate actually increases with catalyst concentration. Catalyst concentration of over approx. 40 millimoles of metal compound per kilograms of the reaction mixture are however disadvantageous for economic reasons. The catalyst is suitably used in an amount that lies between a value slightly above the minimum M

which is calculated using equation (1) and approx. 30 millimoles per kilogram of reaction mixture.

As can be seen from equation (1), the critical catalyst concentration increases with increased water concentration and decreased concentration of p-toluene acid, or expressed differently: oxidation cannot take place in such a water solution if the concentration of p-toluene acid does not exceed a critical value, which increases with reduced catalyst concentration. It is thus not exclusively the catalyst, but also p-toluene acid, which is of significant importance for oxidation to take place in the presence of significant amounts of water. As a result of the combined effect of the catalyst and p-toluene acid, both used in correct amounts according to the present invention, it is actually possible to oxidize p-xylene to terephthalic acid without having to use expensive and/or corrosion-promoting activators, e.g. . bromine compound-ser. This effect of p-toluene acid is a completely unexpected feature of the present invention, as other carboxylic acids, also those of a similar structure, e.g. benzoic acid, does not have the same property.

Furthermore, it appears from equation (1) that the critical minimum concentration of catalyst which is necessary to ensure that oxidation can take place in a homogeneous, water-based system,

also depends on the relative amounts of the manganese and cobalt compounds in the catalyst used. It appears that manganese is markedly more effective than cobalt. For e.g. to allow oxidation to take place in a 50% by weight water solution of p-toluene-

acid (with a molar ratio of 7.56 between water and p-toluene acid), the lowest catalyst concentration needed is 4.5 or 20.9 millimoles/kg, depending on whether manganese or cobalt is used as the only catalyst. For practical reasons, however, it is advantageous to use a mixture of the two metals, as cobalt usually has a favorable effect on the reaction rate. It has also been observed that when a mixture of manganese and cobalt compounds is used as a catalyst, less overoxidation occurs, and that greater yields of terephthalic acid are usually obtained than if each metal is used alone. Thus, e.g. that the combustion ratio (the molar ratio between developed carbon dioxide per absorbed unit of oxygen) usually has a value of between 0.06 and 0.09 in the former case, while values of approx. 0.15 is observed in the second case. The beneficial effect of using manganese and cobalt together as a catalyst is actually a typical feature of the invention and a further advantage of it: when the reaction is carried out in a two-phase system instead of in a homogeneous solution, as is the case in the method according to the invention, one regularly observes combustion ratios of 0.12 or even above this, regardless of whether manganese or cobalt or mixtures of these are used. In most cases, a value of x in equation (1) of between 0.1 and 0.9 is advantageously used to ensure effective oxidation in a homogeneous water-based medium and thereby high yields of terephthalic acid.

For the same practical reasons as stated above, it can also be advantageous if the catalyst, in addition to cobalt and/or manganese, also contains a further metal component, such as nickel. Although this metal is not necessary for oxidation to take place in a homogeneous, water-based medium, they can lead to certain practical improvements with regard to e.g. product purity and reaction rate.

A surprising feature of the invention is that the minimum catalyst concentration to be used to ensure oxidation in a water-based medium can be calculated from equation (1) for an arbitrary composition of the reaction mixture, and that this equation is independent of such an important variable working parameter as the temperature. This is illustrated by the results shown in the subsequent table i, where values of M, obtained experimentally in a large range of varying conditions, are compared with the values calculated by equation (1). Most of these results have been obtained by the following experimental method: In a 1-litre corrosion-resistant autoclave, equipped with a mechanical stirrer, heating jacket, gas inlet pipe, air outlet and dosing pump for injecting liquid, the various substances to be included in the oxidation reaction mixture are filled, i.e. p-toluene acid, water and catalyst. This mixture is then heated with stirring and air is passed through it. As soon as oxidation has begun, a water solution of the catalyst, which has the same catalyst concentration as the original reaction mixture, is successively injected into the mixture. As a result, the reaction mixture is successively diluted with water without changing the catalyst concentration. The starting mixture may also contain p-xylene, provided that its concentration is such that. no phase separation occurs when the reaction mixture is diluted with the water solution of the catalyst.

The degree of oxidation is controlled by continuously measuring the oxygen content of the exhaust gas using analytical oxygen gas meters. As long as the concentration of p-toluene acid in the system is sufficiently high to ensure effective oxidation at the selected catalyst concentration, the reaction proceeds regularly at a rate which decreases successively as a result of the dilution. However, as soon as the critical concentration of p-toluene acid is reached, the reaction rate drops rapidly to near zero. The injection is then interrupted, and the reaction mixture is analysed. The thereby determined molar ratio between water and p-toluene acid is clearly the critical y-value that corresponds to the given catalyst concentration M, or conversely M is the smallest catalyst concentration that can be used to ensure oxidation in a reaction mixture where the molar ratio between water and p-toluene acid is y. In fact, M is still dependent on the catalyst's composition, which can be expressed as the molar ratio between manganese and the combined amount of manganese and cobalt in the catalyst used (x in equation 1).

It is clear that such empirical determinations must be affected by experimentally conditioned errors, so that for a given reaction mixture (for a given y and x value) one can expect to get different values of , which are statistically distributed about the real value with a mean deviation &. The value of M determined in this way for different y^ and x^ values is given in the following table. From this data, it can be estimated that the mean deviation of M, from the value calculated according to equation (1) is 0.70. Consequently, every experimental value which does not deviate from the calculated value by more than two standard deviations, ie by more than 1.4, can be considered to satisfy equation (1).

From the results given in table I, it is clear that: 1. Equation (1) applies to any value of y and x (see e.g. e.g. trials 1 and 7 for y and trials 7 and 19 for x). 2. Equation (1) applies regardless of the presence or absence of p-xylene (see e.g. experiments 8 and 13). 3. Equation (1) applies regardless of temperature (see e.g. experiments 3 and 9). 4. Equation (1) applies regardless of the presence of metals other than manganese and cobalt (see e.g. experiments 2 and 5).

The previously put forward theories relating to the choice of metal compound to be used for oxidation of p-xylene and its partially oxidized derivatives to terephthalic acid are somewhat confusing. As a general rule, however, it is claimed that cobalt is the best catalyst, especially in the absence of a bromine-type activator, and that manganese is less effective, if not completely inactive. For the oxidation of p-toluene acid, manganese has been shown to be inactive (N Ohta et al., "Chem. Abstr." Vol 56, 8620 g. 1962) or even to have an inhibitory character (V N Aleksandrov et al., " Kinet. Catal." Vol 15, 1974, p. 505). Under other conditions, manganese has a catalytic effect, but discoloration of the terephthalic acid occurs if the water content is greater than 10% of the solvent's weight or if the manganese content of the reaction mixture is high. It is therefore highly unexpected that manganese in the method according to the invention has a very strong ability to allow the oxidation reaction to take place in the presence of large amounts of water, and that terephthalic acid is obtained with a high yield in the form of a white, crystalline precipitate, which is particularly suitable for further purification to the degree of purity required for the production of polyester fibers (=fibre quality).

The invention will subsequently be described in more detail with the help of a number of design examples.

Example 1

In a corrosion-resistant autoclave with a volume of one liter and equipped with a mechanical stirrer, heating jacket, gas inlet pipe and vent opening, the following were filled:

The reaction vessel was pressurized with air to a pressure of 20 kp/cm 2 , and the above mixture was heated with stirring and air supply at a flow rate of 92 liters per hour.

In the above mixture, the molar ratio of water to p-toluene acid (y) was (63.0 x 136.15)/(18.02 x 187.5) = 2.54, and the molar

M M M

the ratio Mn/ Mn+ Co in the catalyst (x) was: 1.50/(1.50 + 1.74) = 0.46. Using equation (1), the minimum catalyst concentration that was necessary in this case to ensure oxidation was calculated as follows:

In the present example, the catalyst concentration was actually 10.8 millimol/kg, i.e. approx. three times the minimum amount. In fact, the reaction started spontaneously on heating and proceeded actively. As a result, the temperature rose rapidly and was maintained at 170°C by controlled cooling.

After a reaction time of 180 minutes, the oxygen gas absorption was 40.9 liters (measured at room temperature and a pressure of 1 atm.). The supply of air was then interrupted, and the pressure in the reaction vessel was lowered successively in order to be able to recover unreacted p-xylene by stripping with water. Finally, the reaction vessel was cooled and opened. The precipitate contained therein was filtered, washed with water and dried under vacuum at approx. 80°C. It was then analyzed using a combination of methods, including acidimetric, polarographic and vapor phase chromatographic. It was thereby ascertained that 89.4% of added p-xylene and 23.1% of added p-toluene acid had been converted into the following products:

If one takes into account in the calculation that 4-carboxybenzaldehyde and other intermediate products in a continuous process will be re-circulated and the majority of them would eventually be converted into terephthalic acid, it can be assumed that the yield of terephthalic acid in such a continuous process will be more than 90 mole percent, calculated from the amount of p-xylene consumed.

In another experiment, which was carried out under the same conditions, the hot reaction mixture was filtered. The filter cake thus obtained was washed on the filter with warm water and dried under vacuum. The analysis of the dried product showed that it consisted of (in weight%):

The color of this sample was determined by measuring the optical density of a solution of the sample in dilute ammonia according to a method described in US Patent 3,354,802, with the exception that a 5 cm cell was used instead of a 4 cm cell. The optical density values thus obtained are given in the subsequent Table II, and they were compared with the optical density values obtained with a sample of terephthalic acid of commercial quality and with a degree of purity of 99<+>%.

It appears that the raw sample obtained by simple filtration and superficial washing had better color properties than a terephthalic acid sample of commercial quality with much better purity. This raw sample is therefore particularly suitable for continued purification to fiber quality.

Example 2

The experiment according to the preceding example was repeated, with the exception that instead of 63.0 g of water as solvent, a mixture of:

After 180 minutes of reaction time, the oxygen gas absorption was 3 9.7 litres, i.e. approximately the same amount as in the previous example. The reaction mixture was then processed and analyzed in the same way as described in the previous example. It was thereby ascertained that 88.0% of the added p-xylene and 23.5% of the added p-toluene acid had been converted into the following products:

Using the same method as in the previous example, it can be estimated that the yield of terephthalic acid in a continuous process, which is carried out under the conditions used in the present example, should be approx. 90 mole percent, calculated from p-xylene consumed.

It appears that these results are practically identical to those obtained in the previous example, where water was used as the sole component of the solvent.

Example 3

The experiment in example 2 was repeated, but with the use of formic acid instead of acetic acid. The solvent used in this case was therefore a mixture of:

water 33.0 g

formic acid 3 0.0 g

After a reaction time of 180 minutes, the oxygen gas absorption was 36.7 liters. Using the same analysis method as the previous examples, it was established that 88.0% of the added p-xylene and 15.6% of the added p-toluene acid had been converted into the following products:

Using the same method as in the previous example, it can be estimated that the yield of terephthalic acid in a continuous process under the same working conditions as in the present example will be approx. 8 5 mole percent, calculated from the amount of p-xylene consumed.

Admittedly, these results are not quite as good as the results in the previous examples, but they show that large amounts of formic acid can be allowed in the reaction mixture without any serious negative effects on the yield and reaction rate. This is particularly surprising because formic acid has so far always been described as an effective inhibitor of oxidation reactions. Since formic acid is always formed in such reactions, one has usually been forced to construct elaborate and expensive processes to avoid any accumulation of formic acid in the reaction mixture. In the present process, formic acid is burned, which is confirmed by the fact that the carbon dioxide developed in the present experiment was 6.0 liters instead of 3.8 liters, which was measured in the experiment in example 1. Consequently, in the method according to the present invention, no special operation to remove formic acid from the reaction mixture.

Example 4

The experiment according to example 1 was repeated, but cobalt was used as the only catalyst. The components used were the following:

In this mixture, y was 2.50 and x was clearly 0.00. Consequently, the limit value for cobalt to ensure oxidation was: M 2.50 (0.2<00>) , „ , . ,

M = — 1— — - 6.9 millimoles/kg.

0.0724

The concentration of cobalt acetate in the starting mixture was 11.6 mmol/kg, i.e. approx. twice the limit value. After 180 minutes of reaction time, the oxygen gas absorption was 41.5 litres. The reaction mixture was then processed and analyzed as described above. It was thereby ascertained that 90.8% of added p-xylene and 19.5% of added p-toluene acid had been converted into the following products:

When comparing these results with the results in example 1, it is found that a smaller amount of terephthal

acid despite the fact that the oxygen gas absorption was somewhat greater.

By, as in the previous example, estimating the dividend as

could be achieved in a continuous process, one actually gets a value of 81% instead of 90%. This difference clearly shows the advantage in terms of yield which is achieved by using a mixture of manganese and cobalt as a catalyst instead of cobalt alone.

Example 5

The experiment in example 1 was repeated except that in this case only manganese was used as catalyst. The current composition of the mixture was as follows:

In this mixture, y was 2.60 and x was of course 1.00. Consequently, the limit value of manganese to ensure oxidation was:

In the present example, the manganese concentration was 10.0

mmol/kg, i.e. three times the limit value as in example 1.

After 180 minutes of reaction time, the oxygen gas absorption was 25.3 litres, i.e. significantly lower than in the previous examples. The reaction mixture was then processed and analyzed in the manner described above. It was thereby ascertained that 70.2% of added p-xylene and 5.8% of added p-toluene acid had been converted into the following products:

By comparing these results with the results in examples 1 and 2, it is immediately realized that a smaller amount of reaction participants was consumed in the present case, and that a smaller amount of terephthalic acid was formed. It also applies that the yield of terephthalic acid from a continuous process, calculated on the basis of these data, in the same way as in the previous examples, is only 69%. Despite the fact that manganese is thus significantly more effective than cobalt in ensuring oxidation in the presence of water (which is confirmed by the lower calculated value of M when compared to the value calculated for cobalt alone), a decided advantage is obtained when both reaction rate and yield of the reaction apply to using manganese together with cobalt,

such as in example 1.

Example 6

The experiment in example 1 was repeated, but in this case manganese was used together with 1.5 millimoles of nickel. The actual composition of the mixture was as follows:

In this mixture, y was 2.60 and <x>1.00, as in the previous example. M was thus 3.2 mmol/kg (nickel was not included in the calculation of M). In the present example, the manganese concentration was 5.0 mmol/kg, i.e. 1.8 mmol above the limit value. An active oxidation was actually achieved.

After 180 minutes, the oxygen gas absorption was 29.2 liters, and the reaction mixture was analyzed in the manner already described. It was thereby shown that 79.1% of added p-xylene and 3.3% of added p-toluene acid had been converted into the following products:

By comparing these results with the results from the preceding examples, it is found that somewhat more oxygen gas was absorbed, and that more terephthalic acid was formed than when using only manganese in a concentration twice as large as in the present example. This shows that metals other than cobalt can also be used together with manganese to improve the reaction rate.

Example 7

In the same autoclave that was used in the previous examples, it was filled:

In this mixture, the molar ratio of water to p-toluene acid (y) was 1.04, and the molar fraction of manganese in the catalyst (x) was 0.46. With the help of equation (1), it is thus found that the minimum catalyst concentration that must be used in this case to ensure oxidation is M = 2.7 millimoles/kg. In the present example, the catalyst concentration was more precisely 0.97 mmol for 300 g of the starting mixture, or 3.2 mmol/kg. i.e. only 0.5 mmol greater than the limit value. Nevertheless, oxygen gas absorption started spontaneously and proceeded actively. As a result, the temperature rose rapidly and was maintained at 170°C by controlled cooling.

After 395 minutes of reaction, 62.9 liters of oxygen gas had been absorbed. The air injection was then interrupted, and the same procedure as in example 1 was carried out to determine the composition of the reaction mixture. It was thereby established that 98.8% of added p-xylene and 29.2% of added p-toluene acid had been converted into the following products:

By judging from this data, as in example 1, the yield of terephthalic acid that would be obtained in a continuous process where the intermediate products are recycled, a yield of 80% is obtained, i.e. 10% lower than that determined in example 1. The main difference between the two experiments is that in example 1 the water content of the mixture was 21% by weight, together with only 10%

in the present example. The resulting difference in yield can thus be considered as further evidence of the advantage of carrying out the oxidation of p-xylene in the presence of significant amounts of water according to the invention.

Comparison example

The same amounts of starting material as in the previous example were filled into the autoclave, with the exception that only 0.30 mmol of manganese acetate and 0.35 mmol of cobalt acetate were used. The total concentration of metal catalyst was therefore 2.2 millimol/kg, i.e. 0.5 mmol less than the limit value, as calculated in the previous example. On heating the mixture in the presence of an air stream under the same conditions as in the previous example, the oxygen gas absorption began in the same way. but after approx. 70 minutes it suddenly decreased to an insignificant level. The oxygen gas absorption was only 6.7 liters.

Example 8

In the same autoclave as in example 1, it was filled:

It can be seen that this mixture was very close to the mixtures in examples 1-5, except that p-tolualdehyde was used instead of p-xylene. The molar ratio between water and p-toluene acid was also here 2.54. When applying equation 1, it is therefore found that the lowest catalyst concentration to ensure oxidation is M = 3.2 millimoles/kg. In this example, the catalyst concentration was more precisely 10.8 mmol/kg, i.e. approx. three times as high.

This mixture was heated in the presence of air such as

in the preceding examples. Oxygen gas absorption began spontaneously at approx. 30°C. The temperature was allowed to rise and was kept at 170°C by regulated cooling. After 180 minutes of reaction time, the oxygen gas absorption was 25.6 litres. The introduction of air was then interrupted, and 180 ml of n-heptane was injected into the reaction vessel for the extraction of unreacted p-tolualdehyde, all under continuous stirring and heat supply. The resulting mixture was cooled and the reaction vessel opened. The precipitate that was in this was filtered, washed with n-heptane, dried under vacuum at 50°C, washed again with water and finally dried under vacuum at approx. 70°C. It was then analyzed in the same way as in the previous examples. The filtrate and washings were combined, and the heptane extract was separated by decantation. A sample of this extract was analyzed by vapor phase chromatography to determine unreacted p-tolualdehyde. The remaining portion was evaporated under vacuum to dryness, and the remainder was analyzed according to the same method as the first precipitate.

In these different analyses, it was established that 99.9% of added p-tolualdehyde and 10.7% of added p-toluene acid had been converted into the following products:

Comparison example 1

The same quantities of the compounds as in the previous example were filled into the autoclave with the exception that only

0.3 millimoles of both manganese and cobalt acetate. The total catalyst concentration was therefore 2.0 mmol/kg, i.e. 1.2 mmol lower than the limit value, calculated according to equation (1).

This mixture was heated in the presence of air and under the same conditions as in the previous example. After 180 minutes, the oxygen gas absorption was only 9.0 liters. The reaction mixture was then processed and analyzed in the same manner as

in the preceding example. It was thereby ascertained that the added p-tolualdehyde had been completely converted into the following products:

It is thus found in this case that p-tolualdehyde was mainly converted to p-toluene acid without any significant amount of terephthalic acid being formed.

Comparison example 2

In this example, p-tolualdehyde was used as the only compound to be oxidized. The mixture specifically consisted of the following components:

On heating this mixture in the presence of air under the same conditions as in Example 8, the oxygen gas absorption began in a similar manner, but then proceeded markedly more slowly. After approx. After 240 minutes of reaction at 170°C, the oxygen gas absorption was 29.6 liters. By treating and analyzing the resulting reaction mixture in the manner described above, it was established that 99.7% of added p-tolualdehyde had been converted into the following products:

One also finds here that p-toluene acid was the main product

of the oxidation of p-tolualdehyde, and that only small amounts of terephthalic acid had been formed. This result shows in a striking way that p-toluene acid must be included in a sufficient amount already from the beginning of the reaction when it comes to oxidizing p-tolualdehyde to terephthalic acid according to the present invention.

Example 9

The experiment according to example 8 was repeated with the exception that 3.48 millimoles of cobalt acetate was used as the only catalyst. Using equation (1), it can be calculated that the minimum concentration of cobalt to be used in this case is M = 7.0 mmol/kg. In fact, the concentration of cobalt catalyst in this example was 11.6 mmol/kg.

Here, too, the oxygen gas absorption began at a low temperature. However, heat was supplied in such a way that a temperature of 170°C was maintained. After approx. After 210 minutes of reaction time, the oxygen gas absorption was 20.8 litres. The reaction mixture was then processed and analyzed in the same way as in the previous example. It was thereby ascertained that 99.9% of added p-tolualdehyde and 6.0% of added p-toluene acid had been converted into the following products:

Comparison example

In the same autoclave that was used in the previous examples, the following was introduced:

The same amount of cobalt catalyst was thus used as above (11.6 mmol/kg), but here the molar ratio between water and p-toluene acid was 6.45 instead of 2.54. The minimum concentration of

cobalt catalyst to be used in this case is therefore M = 17.8, i.e. 6.2 mmol more than what is included in the mixture.

This mixture was heated under air supply. Here, too, oxygen gas absorption began immediately. After 45 minutes, however, it dropped to an insignificant level. Nevertheless, the temperature was still kept at 170°C. After 300 minutes of reaction time, the oxygen gas absorption was only 10.0 liters. The reaction mixture was then treated and analyzed in the same way as described in example 8. It was thereby ascertained that 90.1% of added p-tolualdehyde had been converted into the following products:

It is found that the p-tolualdehyde was converted mainly to p-toluene acid without any significant amount of terephthalic acid being formed.

Example 10

In the same autoclave as in the previous example, it was filled:

In this example, neither p-xylene nor p-tolualdehyde was used. The mole ratio (y) between water and p-toluene acid was 2.64, and the mole fraction of manganese in the catalyst (x) was 0.46, as in example 1. The smallest amount of catalyst to be used and which was calculated using equation (1) was M = 3.3 mmol/kg. In the present example, the catalyst concentration was in practice 3.24 mmol per 300 g of reaction mixture, which gives 10.8 mmol/kg, i.e. approx. three times the limit value.

This mixture was heated under air supply under the same conditions as in example 1. Also in this case the reaction started spontaneously. After 180 minutes of reaction time, 19.4 liters of oxygen gas had been absorbed, and the reaction was stopped by cooling. The autoclave was then opened and the sediment in it was treated and analyzed in the same way as described in example 1. It was thereby ascertained that 33.2% of added p-toluene acid had been converted into the following products:

The net yield of terephthalic acid, based on converted p-toluene acid, was 87 mole percent. If, however, one takes the oxidation intermediates, i.a. 4-carboxybenzaldehyde, included in the calculation, the real yield of terephthalic acid that could be achieved in a continuous process where the intermediates are recycled can be estimated to be 97 mole percent.

This example clearly shows that with the method according to the invention p-toluene acid can be oxidized efficiently and with a high yield to terephthalic acid, also in the absence of p-xylene or another easily oxidizable accelerator compound.

Example 11

In the same autoclave as in the previous examples, it was filled:

In this example, the water therefore made up 50% of the starting material mixture's weight. y was 7.95 and x clearly 0.50. According to equation (1), the smallest concentration M catalyst that could be used was 4.9. In the present example, the actual catalyst concentration was 7.0 mmol/kg, i.e. approx. 2.1 mmol higher than the limit value.

Air was admitted into the reaction vessel in a flow of 110 liters per hour and with a pressure of 20 kp/cm 2 , and the mixture was heated with stirring. When the temperature was approx. 170°C, a very small amount of tert-butyl hydroperoxide was added to help the reaction start. As a result, oxygen gas absorption began immediately. The temperature then rose rapidly and was kept at 185°C by regulated cooling. After 300 minutes of reaction, the oxygen gas absorption was 23.5 liters. Then

the reaction was interrupted, and the resulting mixture was treated and analyzed in the same way as in example 1. It was thereby ascertained that 95% of added p-xylene and 4 0.3% of added p-toluene acid had been converted into the following products:

Comparison example 1

Exactly the same experiment as in the preceding example was repeated, with the exception that only half as large an amount of each of the metal catalysts was used. The total catalyst concentration was therefore 3.5 mmol/kg, i.e. approx. 1.4 mmol lower than the limit value.

This mixture was heated under air supply under exactly the same conditions as in the immediately preceding example. No oxygen gas absorption occurred at all when this mixture was held at 185°C for 160 minutes, despite two successive additions of tert-butyl hydroperoxide to help initiate the reaction.

By considering the results obtained in the above examples, where active and significant oxidation took place in the presence of large amounts of water, it is clear that such oxidation can be carried out only using the correct amount of catalyst calculated according to the present invention.

Comparison example 2

The same experiment as in example 11 was repeated with the exception that benzoic acid was used instead of part of the p-toluene acid. The actual composition of the mixture was as follows:

In this example, y = 26.59. With the aid of equation (1), it is thus found that the smallest amount of catalyst that must be used to ensure oxidation of the above-mentioned mixture is 10.7 mmol/kg, i.e. 3.7 mmol more than the amount actually present. If, on the other hand, benzoic acid had the same effect as p-toluene acid in terms of ensuring oxidation in an aqueous medium, this should have been taken into account in the calculation of y and M. In the present case, y = (156.6/ 18.02)/(44.5/136.15 + 93.9/122.12) = 7.93, i.e. the same value as calculated for y in example 11, where active oxidation took place.

The above mixture was heated to and maintained at 185°C under the same conditions as in Example 11, and also in this case a little tert-butyl hydroperoxide was added to facilitate the start of the oxidation. However, no significant oxidation occurred even with continued heating for 180 minutes. This shows that at least benzoic acid does not have the same activity as p-toluene acid to promote oxidation in the presence of water, as occurs in the method according to the invention.

Comparison example 3

The same experiment as in example 11 was repeated with the exception that part of the p-toluene acid was replaced with acetic acid. The actual composition of the mixture was as follows:

In this example, y = 26.92, so that M was 10.8, i.e. approx. the same as in the previous comparison example. Here, too, the acetic acid, if it had the same activating effect as p-toluene acid on oxidation in an aqueous medium, should have been included in the calculation of y and M. In this case, y = (187.2/10.02) would have been obtained /(52.7/136.15 + 55.2/60.05) = 7.95, i.e. exactly the same value as in example 11.

The above mixture was heated to and held at 185°C under the same conditions as in Example 11, and also in this case a little tert-butyl hydrogen peroxide was added to help start the reaction. After a reaction time of 300 minutes, the oxygen gas absorption was only 3.8 liters. This result again shows that the reaction-activating effect of p-toluene acid, as it occurs in the method according to the invention, is not caused by other carboxylic acids, but is consequently specific for p-toluene acid.

Example 12

In the same autoclave as in the previous examples, it was

introduced:

Here y = 4.33 and x = 1.00 with manganese as the only catalyst. Using equation (1), it is found that the smallest manganese concentration that can be used in this case is M = 3.6 mmol/kg. In the present example, the real manganese concentration was 6.0 mmol/kg, i.e. it was well above the limit value

This mixture was heated under air supply under exactly the same conditions as in Example 11, with the exception that the temperature was maintained at 170°C instead of 185°C. After 305 minutes of reaction time, the oxygen gas absorption was 21.2 litres. When treating and analyzing the obtained reaction mixture in the same way as in the previous examples, it was established that approx. 96% of added p-xylene and 29.9% of added p-toluene acid had been converted into the following products:

Comparison example

The same amounts of the same compounds as in the preceding example were introduced into the reaction vessel with the exception that only 0.60 millimoles of manganese acetate were used as catalyst. Its concentration in the mixture was therefore 2.0 millimoles/kg. As a result, no oxygen gas absorption occurred upon heating the mixture for 60 minutes in the presence of air under the same conditions as in the preceding example, even after the addition of tert-butyl hydrogen peroxide to facilitate the initiation of the reaction.

Example 13

In the same autoclave that was used in the previous examples, it was filled:

In this example, cobalt was used as the only catalyst, i.e. x = 0. On the other hand, y = 2.54. Using equation (1), it is found that the lowest cobalt concentration that can be used in this case is M = 7.0 mmol/kg.

In the present example, the actual cobalt concentration was 11.6 mmol/kg.

Air was admitted into the reaction vessel in a stream of

90 liters per hour under a pressure of 20 kp/cm 2 and the mixture was heated to a temperature of 170°C. During heating, the reaction started spontaneously and was allowed to continue for 240 minutes. At the end of the reaction, the amount of absorbed oxygen gas was 29.8 litres. When treating and analyzing the reaction mixture in the manner described in example 1, it was established that 89.3% of the added p-xylene and 37.0% of the added p-toluene acid had been converted into the following products:

Comparison example

The same amounts of the same compounds as in the preceding

example was introduced into a reaction vessel, with the exception that only 0.90 millimoles of cobalt was used as catalyst

acetate. The concentration of this in the mixture was thus 3.0

mmol/kg, i.e. approx. 4.0 mmol lower than the limit value. The result was that no oxygen gas absorption took place when up-

heating this mixture for 240 minutes in the presence of air under the same conditions as in the previous example, even after repeated additions of tert-butyl hydrogen peroxide.

It should be emphasized that the description of the essence of the invention

equal features only relate to certain particularly preferred embodiments

forms. A person skilled in the art will immediately realize that the invention can be varied and modified in many ways within the scope of

the invention. Thus, it applies e.g. that when carrying out the method according to the invention, some foreign substance can be added to the reaction mixture which can be used as an activator in other processes, e.g. an aldehyde or ketone or a bromine-containing compound. Such additions can also

have a favorable effect on e.g. the reaction rate.

Claims (8)

1. Process for the production of terephthalic acid by oxidation tion of a reaction mixture, which contains at least one oxidizable starting material for terephthalic acid in the form of p-toluene acid, mixtures of p-toluene acid and p-xylene and/or partially oxidized p-xylene derivatives and 5-80% by weight of water in the presence of an oxidation catalyst which contains at least one catalytically active metal compound in the form of manganese or cobalt salts of carboxylic acid and possibly an inert organic solvent in an amount of 0-100% of the amount of water, at a temperature of between 140 and 220°C and a pressure that is sufficient to keep at least part of the water in liquid phase and which is between 5 and 40 kg/cm 2 , characterized in that water is used in such an amount that the liquid reaction mixture is practically homogeneous, and that the catalyst is used in an amount that is sufficient to provide a minimum amount of M millimoles of the catalytically active metal compound per kg of the liquid reaction mixture, where M is defined by the following equation: where y is the molar ratio between water and p-toluene acid in the reaction M the mixture, x is the molar ratio Mn , and A = 0.200 + 0.031, <M>Mn <+><M>Co B = 10.9 + 1.3, C = 4.35 + 0.17, D = 0.0724 + 0.0117. 2. Method in accordance with claim 1, characterized in that an amount of water is used in the liquid reaction mixture of between 10 and 60% of the weight of the reaction mixture. 3. Method in accordance with claim 1 or 2, characterized in that acetic acid is used as organic solvent. 4. Method in accordance with one of the preceding claims, characterized in that an amount of oxidation catalyst is used which is sufficient to give an amount of catalytically active compounds that lies between the minimum amount and 40 millimoles per kg of the reaction mixture. 5. Method in accordance with one of claims 1-3, characterized in that an amount of oxidation catalyst is used which is sufficient to give an amount of catalytically active compounds that lies between the minimum amount and 30 millimoles per kg of the reaction mixture. 6. Method in accordance with one of the preceding claims, characterized in that manganese and cobalt salts are used in quantities which lead to the mole ratio x being between 0.1 and 0.9. 7. Method in accordance with one of the preceding claims, characterized in that a catalyst is used which also contains a nickel compound. 8. Method in accordance with one of the preceding claims, characterized in that a reaction mixture is used where the molar ratio between p-toluene acid and p-xylene is between 3 and 15.