Method for Gaseous Phase Nitration of Aromatics

The present invention relates to a method of producing nitrated aromatic compounds, which method involves a gaseous phase nitration. The nitrated aromatic compounds, the nitro-aromatics, are products of importance to the chemical industry, both as such and as intermediates for further processing.

As a rule, nitrations of aromatics have previously been carried out in a liquid phase with nitric acid and with sulphuric acid as a catalyst. Such processes involve that the sulphuric acid, which is a highly corrosive liquid, must be regenerated which, in turn, involves cost increases as regards both work and apparatus. At the nitration of substituted benzenes, there is also the problem of the relation between the different isomers obtained. Thus, at the nitration of toluene, normally approx. 60% ortho, 3% meta, and 37% para-isomers are obtained. As the orthonitrotoluene is the one which has the least value, there is, of course, an interest in methods with more favourable isomer relations.

Processes with solid catalysts or processes which give a greater yield of the isomer desired have previously been described. In these processes, however, either catalysts which have a short life and which are difficult to regenerate, or else catalysts which are highly corrosive, have been used.

Through the present invention, it has now proved to be possible to cope with the above-mentioned drawbacks. The characteristic features of the invention will be noted from the accompanying claims.

According to the invention, heterogeneous catalysts are used, and then particularly acid catalysts of the siliconaluminium type. The aromatic compound in question is conveyed together with nitric acid in a gaseous phase over catalysts, and the reaction seems to take place at the boundary between the gaseous and the solid phase. The catalysts used here are mechanically stable, and easy to regenerate when the activity decreases. They have also been used for a long time in other connections, e.g. in petrochemical processes.

Tests which have been carried out have shown that the activity is high for nitration, and that comparatively small continuous reactors can given large quantities of nitro-aromatics.

The tests also show that a greater reaction is obtained when the catalyst is finely distributed, and this is in all probability due to the fact that the diffusion of the reaction components in the catalyst particles is a speed-limiting factor.

Finally, the tests have shown that an economically more favourable isomer relation is obtained with the present invention than with conventional liquid phase nitration with sulphuric acid as a catalyst.

The invention will now be described in more detail, with reference to the tests accounted for in the following.

Aromatic and 65% nitric acid were gasified individually and conveyed together with nitrogen into a reactor. This, which consists of a glass tube which can be heated, with an inner diameter of 10 mm, was provided with a thermometer pocket and a glass sinter. The catalyst used, which amounted to 10 g, was placed between crushed quartz, and in this way the reaction components had a uniform temperature when they came into contact with the catalyst. The reaction products were conveyed directly down into an alkaline water solution, which prevents any subsequent liquid phase reactions between unreacted starting products. The product mixture was analysed gas chromatographically, and in addition to the aromatic reacted and the nitro-aromatic formed, also the oxidation products benzyl alcohol, benzaldehyde and benzene were determined.

In the above-mentioned table I, the results are shown of tests with different types of catalysts. The mordenite consists of a product from Norton, USA, with the designation Zeolon 200H, stated to have SiO.sub.2 /Al.sub.2 O.sub.3 =10/1. The montmorillonite was obtained from Sud-Chemie AG, Munich, and has the designation K 306 and the formula Al.sub.2 O.sub.3.4SiO.sub.2.H.sub.2 O+XH.sub.2 O. As will be noted from table I, crushed quartz gave practically no reaction whatsoever, while the Mordenite gave a substantial reaction, and the montmorillonite a still better reaction.

In table II, the results of the most active catalyst (Montmorillonite) according to table I are compared when catalyst particles of different sizes are used. As will be noted, a reduction of the particle size from 2-3 mm to 0.2-0.5 mm gives a very substantial increase of the reaction.

Table III shows how the mol relation between nitro-aromatic formed and aromatic added is influenced by the mol relation between added aromatic and added HNO.sub.3. An increase of the proportion of nitric acid gives a considerable increase of the quantity of nitro-aromatic formed.

From table IV it will be noted that if the quantity of aromatic added per liter of nitrogen gas is reduced from 25.4 mmol to 1.8, the yield of nitro-aromatic will decrease, counted on the aromatic reacted from 95 to 72%. There is obviously no reason to reduce the quantity of aromatic added to less than approx. 25 mmol per liter of nitrogen.

In table V, finally, it is shown how the gas speed (measured in liters of nitrogen per g of catalyst and hour) influences the reaction of the aromatic added. The speed of 2.0 gives a considerably greater quantity of nitro-aromatic formed than the other two values.

In all of the tests accounted for above, toluene was used as an aromatic, but similar results have also been obtained with benzene.

The relation between the quantities formed of the different isomers of nitro-toluene has in all the tests using Montmorillonite as a catalyst been approx. 40:5:55 for ortho, meta, and para, respectively, which is considerably more favourable than at conventional nitration of toluene.

In all of the tests described above, the quantity of oxidation products formed, in the form of benzyl alcohol, benzaldehyde and benzene, have amounted to between 0.9 and 1.3% of the quantity of aromatic added, and has not shown any substantial change owing to the reaction conditions applied.

Most of the tests accounted for in the tables shown above have been carried out at 200.degree. C. Minor deviations from this value which have occurred do not seem to have influenced the results.