Chemistry

B. V. SUVOROV, S. R. RAFIKOV, V. S. KULINOVA, and M. I. KHMURA

ON THE MECHANISM OF THE OXIDATIVE TRANSFORMATIONS OF METHYL ALCOHOL, FORMALDEHYDE, AND FORMIC ACID IN THE VAPOR PHASE IN THE PRESENCE OF TIN VANADATE

(Presented by Academician B. A. Arbuzov, October 3, 1956)

In the vapor-phase oxidation of alkylbenzenes on vanadium catalysts, a considerable amount of relatively low-molecular compounds is formed as by-products of the reaction; among these the principal products are formaldehyde, carbon monoxide, and carbon dioxide \((^{1-3})\). In a number of cases, especially in the oxidation of benzene homologues with an isopropyl group \((^4)\), the formation of methyl alcohol, formic acid, and other substances may also be expected. The mechanism of the formation and further transformations of such “fragments” has been insufficiently studied. The available data testify unambiguously only to the fact that, of all the above-mentioned compounds, the least stable under the conditions of vapor-phase oxidation in the presence of heterogeneous catalysts, including vanadium catalysts, are the lower aliphatic alcohols \((^5)\), upon oxidation of which the corresponding aldehydes and products of complete combustion are formed in the greatest amounts, while acids are obtained in insignificant yield. The latter circumstance is attributed to the instability of acids under these conditions \((^6)\).

The purpose of the present investigation was to study the oxidative transformations of methyl alcohol, formaldehyde, and several other oxygen-containing compounds whose formation is possible in the oxidation of alkylbenzenes. The work was carried out in a flow-type apparatus with a metal reaction tube 1100 mm long and 21 mm in diameter. The system for feeding the starting products has been described earlier \((^{3,7})\). The experiments were conducted at a contact time of 0.1–0.2 sec, a feed rate of the starting products of about 50 g per 1 liter of catalyst per hour, and with a considerable excess of air (up to 50 liters, NTP, per 1 g of starting substance). Water was also introduced into the reaction zone in an amount of 350 g per 1 m³ of air. The catalyst was granulated tin vanadate with grains 3–5 mm across.

Starting substances. Methyl alcohol had b.p. 62.5°/694 mm, \(n_D^{20}\) 1.3293, \(d_4^{20}\) 0.791. Formaldehyde was obtained by sublimation of paraform and was used in the form of a 5.3% aqueous solution. Formic acid contained 8.0% water and had b.p. 105°/695 mm, \(n_D^{20}\) 1.3681, \(d_4^{20}\) 1.190. Hydrogen cyanide was introduced in the form of a 6.3% aqueous solution of ammonium cyanide. Carbon monoxide was prepared by decomposition of formic acid and contained 98.0% CO.

Analytical procedure. The gaseous reaction products, upon leaving the reactor, entered a glass scrubber 1.5 m high, irrigated by means of a Patrikeev pump \((^8)\) with a 10% aqueous solution of caustic soda. Formaldehyde, hydrogen cyanide, and carbon dioxide were thereby absorbed by the scrubber liquid, and the exit gas contained only carbon monoxide. Determination of formaldehyde in the scrubber liquid was carried out by iodometric \((^9)\) as well as dimedone \((^{10})\) methods. Hydrogen cyanide was determined by titration with silver nitrate in the presence of potassium iodide according to Deniges’ method \((^{11})\). In separate experiments, after neutralization of the alkaline absorbent with sulfuric acid and subsequent decom-

hydrocyanic acid was isolated from its trap in pure form. Carbon dioxide was determined gravimetrically, from the precipitate of barium carbonate obtained when the scrubber liquid was treated with barium water.

Fig. 1. Oxidation of methanol (A), formaldehyde (B), and formic acid (V) by moist air: I—formaldehyde, II—carbon monoxide, III—carbon dioxide.

To determine carbon monoxide, part of the gas leaving the scrubber, after preliminary washing with sulfuric acid, was passed over hopcalite. The carbon dioxide formed in this process was trapped with barium water.

The results of experiments on the oxidation of methanol showed that it begins to enter into reaction already at a temperature of 310°. The main reaction products were formaldehyde and carbon monoxide (Fig. 1, A). In considering the sequence of their formation, formaldehyde should be assigned to the number of primary intermediate compounds. The formation of carbon monoxide is apparently a consequence of the decomposition of formaldehyde according to scheme (I).

\[ \mathrm{CH_2O} \ \xrightarrow{-\mathrm{H}\cdot}\ \dot{\mathrm{C}}\mathrm{HO} \ \longrightarrow\ \mathrm{CO}+\dot{\mathrm{H}} \tag{I} \]

There are indications in the literature that the HCO radical at a temperature of about 100° completely decomposes into hydrogen and carbon monoxide \(^{(14)}\).

The supposition expressed above is also confirmed by the results of oxidation of formaldehyde itself (Fig. 1, B). As can be seen from Fig. 1, A and B, the courses of the curves characterizing the yields of the products of oxidation of methanol and formaldehyde are very similar to one another. Formic acid was not detected in the reaction products in either case. This fact is explained not so much by the low stability of formic acid under the given conditions as by the fact that already in the early stages of oxidation of the alcohol and the aldehyde, side reactions according to scheme (I) and in other directions are possible.

The experimental data presented in Fig. 1, V, show that in the oxidation of formic acid a considerable amount of carbon dioxide is formed (up to 40%), whereas in the case of oxidation of methanol and formaldehyde its yield does not exceed 10%. Such a difference in the yields of carbon dioxide gives grounds to suppose that formic acid cannot be regarded as an obligatory intermediate product in the complete oxidation of methanol or formaldehyde. Apparently, the reaction proceeds in several directions. Let us also note that the high yield of carbon dioxide in the experiments on the oxidation of formic acid cannot be interpreted as the result of further oxidation of carbon monoxide. Experiments on the oxidation of carbon monoxide under comparable conditions (Fig. 2, B) show that this reaction on tin vanadate at temperatures up to 410° proceeds at a low rate.

On the basis of the experimental data presented above, using the peroxide and chain theories \(^{(12,13)}\), the following general scheme may be proposed for the oxidation of methanol (and formaldehyde):

\[ \begin{aligned} &\mathrm{CH_3OH} \longrightarrow \overset{\mathrm{H}}{\underset{|}{\mathrm{HCO}}} \longrightarrow \mathrm{H\dot{C}O} \longrightarrow \overset{\mathrm{OO}}{\underset{|}{\mathrm{HCO}}} \xrightarrow{+\mathrm{RH}} \overset{\mathrm{OOH}}{\underset{|}{\mathrm{HCO}}} \longrightarrow \\[-0.2em] &\qquad\qquad\qquad \downarrow \qquad\qquad\downarrow \\[-0.2em] &\qquad\qquad \mathrm{CO}+\dot{\mathrm{H}} \qquad \mathrm{OC{-}OOH} \longrightarrow \mathrm{CO_2}+\dot{\mathrm{O}}\mathrm{H} \\[0.4em] &\qquad \longrightarrow \overset{\dot{\mathrm{O}}}{\underset{|}{\mathrm{HCO}}} \xrightarrow{+\mathrm{RH}} \overset{\mathrm{OH}}{\underset{|}{\mathrm{HCO}}} \\[-0.2em] &\qquad\qquad \downarrow \qquad\qquad\downarrow\ \searrow \\[-0.2em] &\qquad\qquad \mathrm{CO_2}+\dot{\mathrm{H}} \qquad \mathrm{H_2}+\mathrm{CO_2}\quad \mathrm{H_2O}+\mathrm{CO} \end{aligned} \tag{II} \]

For the purpose of further verifying the proposed scheme, it seemed of interest to study the oxidation of methanol under comparable conditions, but in the presence of ammonia. Under these conditions one may expect suppression of the reaction forming carbon monoxide and dioxide, owing to the possibility of interaction of the radical $\dot{\mathrm{H}}\mathrm{CO}$ with ammonia with formation of formamide, which subsequently can readily pass into hydrogen cyanide:

\[ \dot{\mathrm{H}}\mathrm{CO}+\mathrm{NH}_3 \xrightarrow{-\mathrm{H}} \mathrm{HCONH}_2 \xrightarrow{-\mathrm{H_2O}} \mathrm{HCN} \tag{III} \]

It was established beforehand that, under the adopted oxidation conditions and when 3–5 g of ammonia were supplied per 1 g of starting product, hydrogen cyanide did not undergo substantial change (Fig. 2, A). It was also shown that, in the temperature range investigated, carbon monoxide does not interact with ammonia (Fig. 2, B), although it is known that at a higher temperature this reaction proceeds with formation of hydrogen cyanide.

The results of experiments on the oxidation of methanol in the presence of ammonia are given in Fig. 2, G. They show that the main direction of the reaction under these conditions is the formation of hydrogen cyanide, the yield of which reached 90%. It is characteristic that, in the interaction of formic acid with ammonia under analogous conditions, the yield of hydrogen cyanide did not exceed 50%. Consequently, the high yield of hydrogen cyanide in the oxidation of methanol (and formaldehyde) cannot be due to intermediate formation of formic acid.

Fig. 2. Oxidation of hydrogen cyanide (A), carbon monoxide (B), formic acid (V), and methyl alcohol (G) by moist air in the presence of ammonia: I—hydrogen cyanide, II—carbon monoxide, III—carbon dioxide.

The results of the last experiments thus confirm the mechanism presented above for the successive transformations of methanol and formaldehyde under conditions of vapor-phase oxidation on a vanadium catalyst.

Institute of Chemical Sciences
Academy of Sciences of the Kazakh SSR

Received
29 IX 1956

CITED LITERATURE

1. L. Ya. Margolis, Usp. khim., 20, 176 (1951).
2. S. L. Sosin, A. M. Sladkov, Usp. khim., 23, 377 (1954).
3. S. R. Rafikov, B. V. Suvorov, A. V. Solomin, Proceedings of the Conference on Catalytic Hydrogenation and Oxidation, Alma-Ata, 1955, p. 241.
4. S. R. Rafikov, B. V. Suvorov, DAN, 82, 415 (1952).
5. L. F. Marek, D. A. Gan, Catalytic Oxidation of Organic Compounds, 1936, pp. 71–164.
6. T. I. Andrianova, S. Z. Roginskii, ZhOKh, 24, 605 (1954).
7. B. V. Suvorov, S. R. Rafikov, Zav. lab., 18, 764 (1952).
8. V. V. Patrikeev, Zav. lab., 13, 1269 (1947).
9. K. Bauyr, Analysis of Organic Compounds, 1953, p. 182.
10. G. Meyer, Analysis and Determination of the Structure of Organic Compounds, 2, 1937, p. 54.
11. G. Faulz, Volumetric Analysis, 1936, p. 159.
12. K. I. Ivanov, Intermediate Products and Intermediate Reactions of the Auto-oxidation of Hydrocarbons, 1949.
13. N. N. Semenov, Some Problems of Chemical Kinetics and Reactivity, 1954.
14. M. Burton, J. Am. Chem. Soc., 60, 212 (1938).
15. P. V. Zimakov, ZhRFKhO, 61, 997 (1929).
16. G. Bredig, H. Elöd, German Patent 550909, Chem. Zbl., 2, 615 (1932).