Chemistry
Unknown
Submitted 1961-01-01 | RussiaRxiv: ru-196101.17874 | Translated from Russian

Abstract

Full Text

Chemistry

R. G. Kostyanovskii

The Reaction of Ethylenimine with Carbonyl Compounds

(Presented by Academician I. L. Knunyants, 18 I 1961)

On the basis of the data of Doughty and co-workers \((^{1})\), the idea has become established in the literature that ethylenimine with aldehydes and ketones under mild conditions gives oxazolidines.

In a previous communication we described the reaction of ethylenimine with formaldehyde \((^{2})\):

\[ \ce{>NH + O=CH2 -> >N-CH2-OH + >N-CH2-N<} \tag{I} \]

Despite the higher polarization of the carbonyl group in formaldehyde as compared with the carbonyl compounds studied in \((^{1})\), we were unable to observe formation of an oxazolidine.

Simultaneously with the publication of our work \((^{2})\), Hiller and Lidak reported that the products of the interaction of ethylenimine with aldehydes, described by Doughty as oxazolidines, correspond in their chemical and physicochemical properties to α-alkyl-N-methylolethylenimines (derivatives of azinimine I \((^{3})\)):

\[ \ce{>N-CH(OH)-R}, \]

where \(R\) is an induction-positive substituent.

In the present work it is shown that this type of interaction also extends to carbonyl compounds with strongly induction-negative substituents (chloral, hexafluoroacetone, diethyl mesoxalate), and thus is general for a whole series of carbonyl compounds. The reactions studied are represented in the scheme:

\[ \ce{[aziridine]NH + O=CH-CCl3 -> [aziridine]N-CH(OH)-CCl3 <-> [aziridine]NH+ -CH(O-)-CCl3} \tag{II} \]

\[ \ce{[aziridine]NH + O=C(CF3)2 -> [aziridine]N-C(OH)(CF3)2 <-> [aziridine]NH+ -C(O-)(CF3)2} \tag{III} \]

\[ \ce{[aziridine]NH + O=C(COOC2H5)2 -> [aziridine]N-C(OH)(COOC2H5)2 <-> [aziridine]NH+ -C(O-)(COOC2H5)2} \tag{IV} \]

α-Trichloromethyl-N-methylolethylenimine (α-trichloromethylazinimine II); α,α-bis-trifluoromethyl-N-methylolethylenimine (α,α-bis-trifluoromethylazinimine III) and α,α-bis-carbethoxy-N-methylolethylenimine (α,α-bis-carbethoxyazinimine IV) are white crystalline substances, readily subliming in vacuum

substance. II, on prolonged storage or heating, turns yellow and decomposes with liberation of chloroform and formation of a viscous resin. The process is evidently due to spontaneous haloform decomposition and polymerization of the N-formylethylenimine formed. III does not change on storage; IV turns yellow and decomposes.

The structure of II and III was proved by identification of the products of haloform decomposition under the action of alkali:

\[ \begin{aligned} \mathrm{II}\quad &\xrightarrow{\mathrm{NaOH}}\ \text{ethylenimine} + \mathrm{CHCl_3} + \mathrm{HCOONa},\\ \mathrm{III}\quad &\xrightarrow{\mathrm{NaOH}}\ \text{ethylenimine} + \mathrm{CHF_3} + \mathrm{CF_3COONa}. \end{aligned} \]

The structure of II, III, and IV was also confirmed by infrared spectra * (Fig. 1), which agree well with N-substituted ethylenimines \((^{2-4})\). In contrast to the previously described methylol derivatives of ethylenimine, in the spectra of II, III, and IV there is a broad band with a maximum at \(2994\ \mathrm{cm}^{-1}\), characteristic of a zwitterion. It is interesting to note that the stability of the adducts of polyhalogenocarbonyl compounds with amines depends substantially on the basicity of the latter. Thus, with amines \((K_a\ 10^{-3}—10^{-4})\) the adducts are unstable and undergo spontaneous haloform decomposition \((^{5-7})\). With ammonia \((K_a\ 1.79—2 \cdot 10^{-5})\) the adducts are stable (chloralammonia), and, finally, with aniline \((K_a\ 5.3 \cdot 10^{-10})\) perfluoroheptanone does not react at all under mild conditions \((^5)\), while hexafluoroacetone under severe conditions gives products of rearrangement of the hexafluoroisopropanol residue into the ring \((^8)\). Therefore the stability of II and III can be explained by the optimal value of the basicity of ethylenimine \((K_a\ 0.81—1.0 \cdot 10^{-6})\). Apparently, for the same reason the literature does not describe adducts of hexafluoroacetone with secondary aliphatic amines whose basicity, despite their structural similarity to ethylenimine, is several orders of magnitude higher (dimethylamine \(K_a\ 5.2 \cdot 10^{-4}\)).

The rate of reaction of ethylenimine with carbonyl compounds increases regularly in the series:

\[ \begin{gathered} \mathrm{O{=}C}< \qquad \mathrm{O{=}CH_2} \qquad \mathrm{O{=}C}< \\ \delta^{+\prime} < \delta^{+\prime\prime} < \delta^{+\prime\prime\prime} \end{gathered} \]

in accordance with the increase in polarization of the carbonyl group. An interesting dependence is observed in the properties of the products of this reaction:

\[ \begin{gathered} \text{I}\quad \text{II}\quad \text{III}\\ \delta_{\mathrm{I}}^{+} < \delta_{\mathrm{II}}^{+} < \delta_{\mathrm{III}}^{+} \end{gathered} \]

The boiling point rises sharply on going from compounds of the first type to the second; both groups of compounds are prone to the ethyleniminomethylation reaction, which is expressed in the formation of methylene-bis-ethylenimines \((^{2,9})\). Compounds of the third group are readily sublimable crystalline products, whose salt-like character increases with increasing electrophilicity of the substituents. In the infrared spectra of compounds of the first and second groups there are bands of vibrations of the C—O and O—H bonds of the hydroxyl group; in the spectra of compounds of the third group there is a distinct zwitterionic region, which indicates localization of the hydroxyl proton on the nitrogen of the amino group.

\[ \text{* I express my gratitude to Yu. N. Sheinker for recording the spectra.} \]

Experimental Part

α-Trichloromethyl-N-methylolethylenimine (II). To a solution of 19.8 g (0.134 mole) of chloral in 20 ml of abs. ether at 0° was added dropwise (30 min) a solution of 5.8 g (0.134 mole) of ethylenimine in 10 ml of abs. ether. After removal of the solvent, the light, syrupy residue crystallized on addition of 5 ml of abs. ether. There were obtained 21.8 g of white crystals, yield 85.1%, b.p. 73–76° at 5–7 mm; after sublimation (50° at 5 mm), m.p. 83.5–84° (with sublimation).

II was also obtained by treating chloral hydrate with ethylenimine in ether (boiling for 1 hr). The product is soluble in ether, alcohol, benzene, chloroform, and dioxane; sparingly soluble in water; insoluble in petroleum ether. It crystallizes from ether, dibutyl ether, and kerosene. Mol. wt. found 198.8, 204.6; calculated 190.47.

Fig. 1. Infrared spectra of α-trichloromethyl-N-methylolethylenimine and α,α-bis-trifluoromethyl-N-methylolethylenimine.

Found, %: C 25.60; 25.61; H 3.36; 3.55; N 7.50; 7.57; Cl 55.71; 55.79
C₄H₆ONCl₃. Calculated, %: C 25.22; H 3.18; N 7.35; Cl 55.84

α,α-Bis-trifluoromethyl-N-methylolethylenimine (III). To a solution of 7 ml of hexafluoroacetone in 30 ml of abs. ether at −60° was added dropwise (1 hr) a solution of 3.2 g of ethylenimine in 30 ml of abs. ether. The mixture was then brought to room temperature, boiled for 5 min, and left overnight at room temperature. After removal of the solvent, 9.2 g of a white crystalline product was obtained, yield 59.3%. After sublimation (100° at 20 mm), m.p. 147–149.5°.

The product is soluble in ether, acetone, and dioxane; poorly soluble in chloroform and insoluble in water, benzene, and petroleum ether.

Found, %: C 28.29; 28.35; H 2.22; 2.26; N 6.86; 6.93; F 53.74; 53.93
C₅H₅ONF₆. Calculated, %: C 28.82; H 2.41; N 6.7; F 54.51

α,α-Bis-carbethoxy-N-methylolethylenimine (IV). A solution of 9.85 g of diethyl mesoxalate in 50 ml of abs. ether at 0° was treated with a solution of 2.5 g of ethylenimine in 30 ml of abs. ether. The mixture was left overnight with cooling; large colorless plates separated. Obtained 9.9 g, yield 80.6%, m.p. 68–69° (from ether).

Found, %: C 49.82; 49.98; H 7.10; 7.19; N 6.50; 6.55
C₉H₁₅O₅N. Calculated, %: C 49.76; H 6.96; N 6.44

Received
10 I 1961

References Cited

  1. J. B. Doughty, C. L. Lazzell, A. R. Collet, J. Am. Chem. Soc., 72, 2866 (1950).
  2. R. G. Kostyanovskii, DAN, 135, No. 4, 853 (1960).
  3. S. A. Giller, M. Yu. Lidak, Abstracts of Reports at the Symposium on the Chemistry of Antitumor Substances, December 1–2, 1960, Moscow, 1960.
  4. Yu. N. Sheinker, E. M. Peresleni, G. I. Braz, ZhFKh, 29, 518 (1955).
  5. M. Hauptschein, R. Robert, J. Am. Chem. Soc., 77, 4930 (1955).
  6. H. E. Simmons, D. W. Wiley, J. Am. Chem. Soc., 82, 2288 (1960).
  7. A. W. Hofmann, Ber., 5, 247 (1872).
  8. I. L. Knunyants, Chen Chin-yun, N. P. Gambaryan, E. M. Rokhlin, Journal of the All-Union Chemical Society named after Mendeleev, 5, No. 1, 114 (1960).
  9. A. Dornow, W. Schacht, Ber., 82, 464 (1949).

Submission history

Chemistry