I. V. Gebler, Professor at STI, and student M. Tutunov.

On the Question of the Catalytic Oxidation of Toluene in the Vapor Phase

Among the many methods of obtaining benzaldehyde, the possibility of producing it by the direct oxidation of toluene with atmospheric oxygen in the presence of catalysts is of great interest, since here, with a favorable course of the process, one may expect very advantageous operating conditions. Meanwhile this reaction, which in many cases has found such extensive technical application, has been very little developed with regard to the question addressed here—the oxidation of toluene.

Prof. N. N. Vorozhtsov, in his book Fundamentals of the Synthesis of Dyes, devoting considerable space to the description of various methods of obtaining benzaldehyde, notes the insufficient development of the question of the direct oxidation of hydrocarbons with alkyl groups in the nucleus in general, and says that much work still remains to be done in this direction. Indeed, there is not much material in the literature on this question. Thus in Beilstein’s¹ we find: “upon heating toluene with NiO on a water bath for many hours at \(T = 100^\circ\), benzaldehyde and benzoic acid are obtained...” “For the oxidation of toluene, cerium salts or cerium dioxide with 60% \(H_2SO_4\) are also used; at \(T = 90^\circ\) benzaldehyde, a small quantity of phenyl-tolyl-methane, and anthraquinone are obtained.” According to Konovalov, upon heating to \(130^\circ\) in tubes toluene and \(HNO_3\) (sp. gr. 1.07) for 5 hours, benzoic acid and a little oil are separated, which distills between \(109\text{–}111^\circ\) and has the odor of benzaldehyde.

As for the catalytic oxidation of toluene in the vapor phase proper, on this subject there is the statement by Loew¹¹, who observed the formation of benzaldehyde when a mixture of toluene vapors with air was passed over heated CuO. Further, Beilstein¹¹¹, on the basis of the patent literature, says that benzaldehyde is obtained when toluene vapors and air are passed over incandescent platinum wire, and also that when the same mixture is passed through a tube containing heated asbestos with ammonium vanadate salt, aldehyde, benzoic acid, and anthraquinone are obtained. German patents Nos. 168291, 203848, and 208348 also indicate the possibility of obtaining benzaldehyde by passing a mixture of toluene vapors and air (sometimes also together with steam) over pumice on which oxides of Cu, Fe, Ni have been deposited, at \(T\) about \(150^\circ\), and also over coal and peat at \(T: 150\text{–}300^\circ\) (?).

In Ch. Zentralblatt (1923, vol. II, p. 745) a more detailed description is given of English patents Nos. 189081 and 189107 (1922). It is stated here that “aromatic hydrocarbons with alkyl groups in the nucleus, in the vapor state, with oxygen-containing gases, are passed over catalysts consisting of one or several oxides of metals of groups V or VI of the periodic system, with the exception, however, of vanadium oxide; here air, pure oxygen, or ozone, or a mixture of these gases, may be used, with or without the addition

¹ Beilstein, Handb. d. Org. Chemie, vol. V, p. 284.
¹¹ J. prakt. 33, p. 324.
¹¹¹ Beilstein, Handb., vol. VII, p. 174.

diluent gas or water vapor; for preparing the catalyst, pumice or asbestos is taken and impregnated with solutions of salts, especially organic salts of the indicated metals, or with a suspension of finely pulverized metal oxide or oxide hydrate, after which it is ignited. The method is suitable for obtaining \(C_6H_5COH\) by oxidation of toluene, using one of the oxides of U or Mo with or without the addition of CuO to them.

As is evident from the cited reference, the literature contains no sufficiently detailed indications on the conduct of the process of catalytic oxidation of toluene; nothing at all is said about the yields of benzaldehyde, nor about the influence of the nature of the catalysts on the course of the oxidation reaction. In order to shed some light on these questions, we carried out several experiments on the oxidation of toluene with atmospheric oxygen, using various catalysts.

The experiments were performed as follows. As the substrate on which the substances serving as catalysts were deposited, we took pumice ground into grains approximately 0.5 mm in diameter. The pumice was impregnated with a solution of the corresponding salt of one or another metal, dried, and in this form introduced into a refractory glass tube about 2 cm in diameter, which was placed in an ordinary furnace for organic analysis. By ignition in a current of air the salt taken was converted into the corresponding oxide (in the case of Pt, into the free metal), which was thus deposited on the large surface of the ground pumice, constituting as a whole the contact in our experiments; it was taken in such an amount that, with the indicated tube diameter, the length of the contact layer was for the most part about 20 cm. We impregnated the pumice with solutions of the following salts (concentration 2–4 g per liter).

1) Uranyl nitrate \(UO_2(NO_3)_2\), converted on ignition up to \(300^\circ\) into \(UO_3\); at a higher temperature—into the oxide-oxide \(UO_2 \cdot 2UO_3\).

2) Ammonium molybdate \((NH_4)_2MoO_4\); on ignition—\(MoO_3\).

3) Cobalt acetate: on ignition—\(Co_2O_3\).

4) Ammonium vanadate \(NH_4VdO_3\); on ignition—\(Vd_2O_5\).

5) Chloroplatinic acid \(PtCl_4\); on ignition after wetting with methyl alcohol, it was deposited in the form of platinum black.

In addition to all this, copper oxide was also tested as a catalyst; for this purpose, a copper gauze rolled into a sausage 10 cm long was introduced into the reaction tube.

The reaction tube was connected on one side with a condenser and receiver, and on the other with a flask into which toluene was placed; air was introduced into the flask through a tube immersed in the toluene, its consumption being measured by its outflow from a graduated gasometer. By maintaining one or another temperature in the flask, it was possible to regulate the degree of its evaporation and, consequently, the amount of vapor in the mixture with air entering the reaction space; at the same time, the time during which a definite amount of toluene and air passed through the contact was also recorded. The amount of benzaldehyde in the reaction products was determined by the method of K. Ripper, i.e., by treating the alcoholic solution of \(C_6H_5COH\) with titrated \(KHSO_3\) solution and subsequent titration with iodine solution. In the following tables the most characteristic figures from a large number of experiments are given; in the tables the amount of air is given in liters per

1 g of toluene taken into the reaction, while the yield of benzaldehyde is expressed in % relative to the same toluene; the rate of passage of the toluene vapors is expressed as the number of grams thereof in 1 hour, and the rate of passage of air as the number of liters in 1 hour.

1. Experiment with \( \mathrm{UO_3} \).

At a higher temperature, above \(400^\circ\), when \( \mathrm{UO_3} \) passes into \( \mathrm{UO_2 \cdot 2UO_3} \), the formation of benzaldehyde was not observed at all; consequently, uranium protoxide-oxide is inactive in this respect.

2. Experiment with \( \mathrm{MoO_3} \).

Below \(400^\circ\), in this case, formation of benzaldehyde was not observed. The best yield, as may be seen, is obtained at the lowest rate of passage of the toluene vapors with air and with a not very large excess of the latter.

3. Experiment with Pt.

The use of platinum black as a catalyst gives higher yields, and moreover with a smaller excess of air than was the case when using \(\mathrm{MoO_3}\).

4. Experiment with \(\mathrm{Vd_2O_5}\).

Experiments with vanadium showed that \(\mathrm{Vd_2O_5}\), obtained by direct decomposition of \(\mathrm{NH_4VdO_3}\), is in itself inactive, but when toluene vapors are passed over it, it is reduced to \(\mathrm{Vd_2O_3}\), which proves to be a very energetic catalyst, making it possible to obtain up to 7% benzaldehyde under conditions of slow flow of toluene vapors with air, with a very large excess of the latter.

5. Experiment with \(\mathrm{CuO}\).

In this case the greatest yield of benzaldehyde was obtained with a very slow current of hydrocarbon vapors and air.

6. Experiment with \(\mathrm{Co_2O_3}\).

No yield of benzaldehyde was observed at all.

In all our experiments, while adhering to the indicated temperatures, we did not observe any appreciable complete combustion of the oxidized substance, since no \(\mathrm{CO_2}\) was found in the reaction products. Thus, under the given conditions only incomplete oxidation of toluene takes place, with benzaldehyde being formed predominantly; traces of benzoic acid were established only in individual cases; the bulk of the toluene passes through unchanged.

Turning to an evaluation of the results obtained, it should be noted that our figures cannot be regarded as final, and it is quite possible that, under conditions of a similar regime for this process, higher aldehyde yields may be achieved by using more perfect apparatus, which would make it possible to observe more strictly a number of conditions, such as maintaining the temperature within narrower limits, ensuring more intimate mixing of the hydrocarbon vapors with air, and so forth, which promotes the success of catalytic reactions in the vapor phase.