Application of the Formaldehyde Method for the Separation of Carbazole and Anthracene from Crude Anthracene

L. P. Kulev, V. P. Lopatinskii

The separation from crude anthracene of the three principal components—anthracene, carbazole, and phenanthrene—cannot be considered satisfactorily solved up to the present time. Dozens of methods for processing crude anthracene have been proposed, but most of them are of little effectiveness. Only two methods have found industrial application: the potassium-melt method and the pyridine method [1, 2]. In both methods, the separation of anthracene and carbazole is preceded by removal of the bulk of the phenanthrene and oils by washing with such solvents in which anthracene and carbazole are poorly soluble. However, all variants of the potassium-melt method [1, 2, 3, 4, 5, 6], including the recently proposed method for removing carbazole by treating enriched anthracene with potassium alcoholates or phenolates [7, 8, 9, 43], are very labor-intensive and energy-intensive. The pyridine method is also not free of shortcomings [1, 2, 6, 10—16, 43, 44]. Considerable difficulties, in particular, are created by the toxicity and scarcity of pyridine bases in organizing large-tonnage production of carbazole. Recently, other methods for separating anthracene and carbazole have also been developed, for example: azeotropic distillation of crude anthracene with ethylene glycol [16, 17], diethylene glycol [18], or octanol [19]; extraction of carbazole with cyclohexylamine [20], furan derivatives [21, 22], and acetone [23, 24]; extraction of anthracene with the methyl ether of diethylene glycol [25] or with liquid anhydrous hydrogen fluoride [26]. Sulfuric-acid methods for separating crude anthracene [27—32] have continued to be improved, as has the rectification method [37, 45, 46]. Of these, only the last two can have industrial significance, although, as the studies of V. V. Svyatoslavskii [43] show, complete separation of carbazole from the other components cannot be achieved by rectification and, apparently, the additional introduction into this method of chemical processes for isolating and purifying carbazole is required.

Among the many methods for separating carbazole and anthracene, the formaldehyde method deserves attention; it is based on using, for the isolation of carbazole, the reaction of its interaction with formaldehyde, which leads to the formation of 9-oxymethylcarbazole. Anthracene and the other components of crude anthracene do not react with formaldehyde. 9-Oxymethyl-

\[ \begin{array}{c} \text{carbazole} + \mathrm{CH_2O} \longrightarrow \text{9-oxymethylcarbazole} \end{array} \]

\[ \begin{array}{ccc} \text{carbazole structure with } \mathrm{NH} & +\ \mathrm{CH_2O}\ \longrightarrow & \text{carbazole structure with } \mathrm{N{-}CH_2OH} \end{array} \]

Carbazole dissolves well in most organic solvents, owing to which it can easily be separated from anthracene. Carbazole is regenerated from 9-oxymethylcarbazole by boiling the latter in aqueous alkali solutions or by treatment with steam. The possibility of separating carbazole and anthracene by this method was indicated in a patent by Mikeska [33]. Studies of the formaldehyde process for extracting carbazole from crude anthracene were carried out by B. V. Maksorov [34] and Yura and Oda [35]; however, many data necessary for carrying out the indicated process under industrial conditions are lacking in these works.

The aim of the present work was a systematic study of the influence of the principal factors on the yield of 9-oxymethylcarbazole, in order to establish the optimum conditions for isolating carbazole from crude anthracene, as well as the conditions for regenerating formaldehyde. Samples of pure carbazole, artificial mixtures of it with anthracene, and crude anthracene from the Kemerovo Coke-Chemical Plant were studied. In this connection the influence of alkaline additives, their amount, the nature of the solvent, the duration of heating of the reaction mixture, and the duration of holding it after heating were determined. When working with crude anthracene, the main quantities of phenanthrene and oils were first removed from it, since under the conditions of the formaldehyde process the latter passed into solution together with 9-oxymethylcarbazole and contaminated the isolated carbazole.

The carbazole content was determined by the nitroso method of V. I. Khmelevskii and I. S. Levin [38, 39], and in a number of cases directly by the weight of 9-nitrosocarbazole or by the method of M. A. Il’inskii and R. B. Roshal’ [36]. Analyses for anthracene were carried out by the method of I. Ya. Postovskii and V. I. Khmelevskii [39, 40, 41] using maleic anhydride, and the phenanthrene content was established by the method of V. I. Khmelevskii and I. S. Levin [39, 40, 41].

In studying the influence of alkaline additives on the yield of 9-oxymethylcarbazole it was found that formation of the latter is possible even in the absence of alkali (the experiments were carried out in quartz vessels), but only to a very slight extent. The introduction of alkalis, however, substantially accelerates the reaction of formation of 9-oxymethylcarbazole. In order to establish the optimum amounts of alkaline additives, we carried out experiments with various substances.

The results of some of these experiments are given in Table 1.

Table 1

The data presented (Table 1) show that, as the amount of alkali added is increased, the yields of 9-oxymethylcarbazole rise to a certain maximum and then begin to decrease. The latter is explained by the fact that, in the presence of large amounts of alkali, more complex condensation products are formed. In some experiments it was possible to isolate a small amount of amorphous substances with m.p. 290°C, from which carbazole is not regenerated. The maximum yields of 9-oxymethylcarbazole (above 90%) were obtained in the presence of 15–30% alkali (in addition to lime), based on the weight of carbazole. In these experiments it was found that high yields of 9-oxymethylcarbazole can be achieved in the presence of lime, with additions of 3–10% lime being optimal.

It was also established that holding the reaction mixture after heating, as recommended by Makorov [34], is not necessary. The results obtained in experiments with lime were refined in the separation of carbazole from artificial mixtures of it with anthracene and from enriched anthracene. The data from these experiments are given in Table 2.

Table 2

Thus, a high yield of 9-oxymethylcarbazole can be achieved in the presence of lime when separating carbazole from mixtures of it with anthracene and from enriched anthracene. At the same time, 80–85% anthracene is obtained, with a yield of 90–95% of its content in the enriched anthracene.

The experiments showed that the duration of boiling in alcohol when extracting carbazole from enriched anthracene is 30–45 min. Further heating leads to a decrease in the yield of 9-oxymethylcarbazole and to the formation of complex condensation products.

It should be noted that, in order to obtain high yields of 9-oxymethylcarbazole from enriched anthracene, a considerable excess of formaldehyde is required (on average 5 mol per mol of carbazole, Table 3).

Table 3

Of the solvents tested, the best proved to be the lower aliphatic alcohols—ethyl and methyl alcohols; the latter has economic advantages in organizing the industrial isolation of carbazole by this method. The optimum amounts of solvent are up to 30 moles per mole of carbazole.

Application of the formaldehyde method to crude anthracene containing large amounts of phenanthrene and oils did not lead to satisfactory results. A check of methods for the preliminary removal of phenanthrene and oils showed that, for use of the formaldehyde method, anthracene enriched by recrystallization from coal-tar solvents is suitable. The optimum amount of solvent relative to the weight of crude anthracene from the Kemerovo Coke-Chemical Plant should be 120–150%. Washing out phenanthrene was carried out at 80°C. After the solution was cooled to 20°C, the precipitate that separated contained 5–10% phenanthrene and other impurities to anthracene and carbazole (the precipitate of enriched anthracene contains on average 32–36% anthracene and 57–60% carbazole). Complete removal of these impurities, which dissolve in alcohols and contaminate 9-oxymethylcarbazole, cannot be achieved either by treatment with hot water or by steam distillation. Further improvement in the quality of the isolated carbazole can be achieved by reducing the oil content in the crude anthracene.

Regeneration of carbazole from its 9-oxymethyl derivative is readily carried out by boiling the latter with concentrated aqueous ammonia solution for 30 min. The carbazole obtained has a purity in the range of 93–97%. Simultaneously with the regeneration of carbazole, formaldehyde (liberated upon decomposition of 9-oxymethylcarbazole) is also utilized in the form of urotropine, the purity of which reaches 97%. During the isolation of urotropine, the solvent can also be distilled off.

On the basis of the enlarged laboratory experiments we have carried out, data have been obtained that may be used as the basis for developing a technological process for isolating carbazole and anthracene from crude anthracene. The consumption of the main raw materials per 1 metric ton of enriched anthracene is: alkali (lime or caustic soda) 25–60 kg, formaldehyde (40%) 1.5 metric tons, solvent 3.45 metric tons; 95–97% of the formaldehyde is utilized in the form of urotropine, and the solvent is in circulation.

The formaldehyde method for isolating carbazole is readily practicable, gives good results with respect to the quality and completeness of separation of anthracene and carbazole, and may be recommended for industrial testing. It can also be used as one of the final stages in obtaining pure carbazole by any other methods.

Conclusions

1. The influence of various alkaline catalysts on the yield of 9-oxymethylcarbazole was studied. It was found that, in the presence of lime (3–10% of the weight of carbazole), high yields (90–95%) of 9-oxymethylcarbazole can be obtained.

2. Experiments were carried out on the isolation of carbazole by the formaldehyde method from artificial mixtures and from preliminarily enriched crude anthracene. Under optimum conditions, high yields of 9-oxymethylcarbazole, up to 84–87%, can be achieved, and at the same time 80–85% anthracene is obtained with a yield of up to 90–95% of its content in the enriched anthracene.

3. It was established that, upon decomposition of 9-oxymethylcarbazole with ammonia, all formaldehyde can be utilized in the form of urotropine, the purity of which after isolation from the mother liquor reaches 97%.

References

1. Belov K. A., Processing of Chemical Products of Coking. Metallurgizdat, Kh.-M., 1949.
2. Nilyuskii N. A., Crude Anthracene of Coal Tar and Methods of Its Processing, GONTI, Kh., 1938.
3. Tsukerman I. I., Daeva-Stepanenko I., POKh, 3, 143, 1937.
4. Tsukerman I. I. POKh, 5, 248, 1938.
5. Afremov V. I. AKP, 1, 9, 1931.
6. Bojanowski I., Seniow S., Przem. Chem., 4, 285, 1948; C. A., 43, 209, 1948.
7. Insinger T. Am. pat. 2464811; C. A. 43, 4450, 1949.
8. Swietoslawski W. Am. pat. 2464833; C. A. 43, 4451, 1949.
9. Gorzynska J. Przem. Chem. 10, 411, 1954.
10. Nilyugin G. T. ZhPKh, 5, 1059, 1932.
11. Gulinov V. G. Coke and Chemistry, 10, 16, 1933.
12. Il’inskii M. A., Zalkind A. A., Bratsburg L. I. AKP, 4, 152, 1934.
13. Gluzman L. D. Proceedings of UKhIN, 5—6, 104, 1937.
14. Grigor’ev S. M., Khadzhinov V. N. POKh, 4, 345, 1937; Proceedings of UKhIN, 8, 199, 1937.
15. Clark L. M. Ind. Eng. Chem., 11, 204, 1919.
16. Bellet E., C. A., 48, 4809, 1954.
17. Saunier A., Fr. pat. 976773; C. A. 47, 2464, 1953.
18. Feldman J. Orchin M. Amer. pat. 2590096; C. A. 46, 10208, 1952.
19. Andrews J. Am. pat. 2675345; C. A. 48, 9658, 1954.
20. Hatton J., May R., Engl. pat. 611554; C. A. 43, 2641, 1949.
21. Costabel L. D. It. pat. 444330; C. A. 45, 3148, 1951.
22. Ashida S., Shimizu Y., Coal Tar 4, 357, 1952; C. A. 47, 6119, 1953.
23. Dutch pat. 74130; C. A. 48, 12395, 1954.
24. Stamicarbon N. V. Engl. pat. 726860; C. A. 49, 11264, 1955.
25. Kleiss L., Marwil S. Amer. pat. 2715647; C. A. 49, 15216, 1955.
26. Lien A. P. Amer. pat. 2426624; C. A. 42, 223, 1948.
27. Borodkin V. F. ZhPKh, 23, 763, 1950.
28. Venkataraman K. Chemistry of Synthetic Dyes, vol. 1, GKhI, L., 1956.
29. Rottschaefer B. Amer. pat. 2459135; C. A. 43, 3465, 1949.
30. Dutch pat. 67753; C. A. 45, 9084, 1951.
31. Dutch pat. 68993; C. A. 46, 9608, 1952.
32. de Bruijn H. Amer. pat. 2575314; C. A. 46, 8152, 1952.
33. Michael I. Germ. pat. 386597; Friedländer, 14, 544, 1926.
34. Maksorov B. V. ZhKhP, 6, 41, 1929.
35. Jura S., Oda R. J. Soc. Chem. Ind. Jap., 43, 392, 1940; C. A. 35, 4019, 1941.
36. Il’inskii M. A., Roshal’ R. B. DAN SSSR, New series, 17, 117, 1937.
37. Tsypkina O. Ya. ZhPKh, 28, 185, 1955.
38. Khmelevskii V. I., Levin I. S. POKh, 17, 308, 1940.
39. Khmelevskii V. I., Postovskii I. Ya. ZhPKh, 17, 463, 1944.
40. Postovskii I. Ya., Khmelevskii V. I. ZhPKh, 10, 759, 1937; Proceedings of the Ural Industrial Institute, 6, 64, 1938.
41. Lastovskii R. P. Technical Analysis in the Production of Intermediates and Dyes, 2nd ed., GKhI, M.-L., 1949.
42. Khmelevskii V. I., Levin I. S. POKh, 7, 241, 1940.
43. Sventoslavskii V. V. Physical Chemistry of Coal Tar, Izd.-inlit, M, 1958.
44. Gluzman L. D. Coke and Chemistry, 4, 35, 1958.
45. Goftman M. V., Emel’yanova V. P. Proceedings of the Ural Polytechnic Institute, 59, 47, 1957.
46. Isaenko I. P., Markacheva G. M. Coke and Chemistry, 12, 35, 1958.

Corrections and Noted Misprints

Izv. TPI, vol. 102.