CHEMISTRY

Ya. F. Freimanis and Academician of the Academy of Sciences of the Latvian SSR G. Ya. Vanag

Salts of 3-Aminoindones

Previously we obtained the hydroiodides of 3-alkylamino-2-phenylindones, which represent a little-studied class of substances: salts of β-aminovinylcarbonyl compounds. Such salts are of interest from the theoretical point of view as salts of vinylogs of amides and therefore have been studied in greater detail. Salts of β-aminovinyl ketones with acids have been obtained either by an indirect route \((^{1-6})\), or directly from the base and the corresponding acid \((^{7-9})\). The existence of salts with two equivalents of acid has even been proposed (though in conc. \(\mathrm{H_2SO_4}\)), which, however, cannot be isolated \((^{10})\). It may be expected that 3-aminoindones, like other vinylogs of amides, will prove to be very weak bases. Salts of 3-aminoindones with acids are obtained directly from the corresponding aminoindone (I) and a hydrogen halide. The hydroiodides are the most stable, whereas HCl hydrolyzes most aminoindones under the action of traces of water to 2-substituted indanediones (see below). The hydrohalides are obtained by passing a stream of dry hydrogen halide through a solution or suspension of aminoindone (I) in a suitable solvent \((\mathrm{CH_3COOH},\ \mathrm{CH_2ClCH_2Cl},\) tetrahydrofuran (THF)). All the hydrohalides are unstable substances, losing hydrogen halide under the action of moisture, on heating, and often under vacuum. Therefore the analytical data do not always agree exactly with the calculated values. It is likewise pointless to determine their melting points. For the hydrohalides of aminoindones, structures (II) or (III) may be proposed.

\[ \begin{array}{cccc} \begin{array}{c} \mathrm{NHR_1}\\ |\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CR_2}\\ \backslash \ /\\ \mathrm{CO} \end{array} & \begin{array}{c} {}^{\oplus}\mathrm{NHR_1}\quad \mathrm{J}^{\ominus}\\ \|\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CHR_2}\\ \backslash \ /\\ \mathrm{CO} \end{array} & \begin{array}{c} {}^{\oplus}\mathrm{NHR_1}\quad \mathrm{Cl}^{\ominus}\\ \|\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CR_2}\\ \backslash \ //\\ \mathrm{C}\\ |\\ \mathrm{OH} \end{array} & \begin{array}{c} \mathrm{NHR_1}\quad \mathrm{Cl}^{\ominus}\\ |\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CR_2}\\ \backslash \ /\\ \mathrm{C}\\ \|\\ {}^{\oplus}\mathrm{OH} \end{array} \\[1.5em] \mathrm{(I)} & \mathrm{(II)} & \mathrm{(IIIa)} & \mathrm{(IIIb)} \end{array} \]

\[ \begin{array}{ccc} \begin{array}{c} {}^{\oplus}\mathrm{NH_2}\quad \mathrm{J}^{\ominus}\\ \|\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{C}\\ \backslash \ / \ \backslash\\ \mathrm{CO}\quad \mathrm{R_2}\\ \quad\quad /\\ \quad\mathrm{R_1} \end{array} & \begin{array}{c} \mathrm{CO}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CHR}\\ \backslash \ /\\ \mathrm{CO} \end{array} & \begin{array}{c} {}^{\oplus}\mathrm{NH_2}\cdot 2\mathrm{HSO_4}^{\ominus}\\ \|\\ \mathrm{C}\\ / \ \backslash\\ \mathrm{C_6H_4}\quad \mathrm{CHC_6H_5}\\ \backslash \ /\\ \mathrm{C}\\ \|\\ {}^{\oplus}\mathrm{OH} \end{array} \\[1.5em] \mathrm{(IV)} & \mathrm{(V)} & \mathrm{(VI)} \end{array} \]

The UV absorption spectra of the hydroiodides of 3-alkyl- and 3-arylaminoindones and of the hydroiodides of imines of 2-phenyl- or 2-methylindandione are similar to the spectra of fixed hydroiodides of 2,2-disubstituted indandiones (IV) (Tables 1 and 2; in Table 2, previously obtained substances are also included for comparison). Consequently, the hydroiodides have structure (II); they are colored in ...

Table 1

Obtained hydroiodides of aminoindones

yellow or orange-yellow in color, whereas the starting aminoindones are red or orange substances. In the IR spectrum (in solutions), salts (II) have a band \(\nu_{C=N<}^{\oplus}\) and a strongly enhanced band \(\nu_{C=O}\) (Table 2). In solutions the overall absorption intensity is low, since the hydroiodides are only very sparingly soluble in \(\mathrm{CH_2ClCH_2Cl}\). For solid (II), the aforementioned bands are noticeably lowered. 2-Mono- and 2,2-disubstituted indandiones (V) of diketone structure have two carbonyl bands—on average at 1710 cm\(^{-1}\) and 1740 cm\(^{-1}\) (the less intense one)—as a result of interaction of the vibrations of the two carbonyl groups \(^{(11-15)}\). In our case it is of interest that, when one ketone group is replaced by an ammonium group, it is precisely the higher carbonyl band that remains. It is possible that in this case too the stretching vibrations of the groups \(\mathrm{C=O}\) and \(\mathrm{C=N<}^{\oplus}\) are linked to some extent. Our data on the absorption of salts (II) agree with the work of other authors on the IR spectra of β-ketoimmonium compounds \(^{(9,16)}\).

It was possible to obtain hydrochlorides of the imines of 2-methyl- and 2-phenylindandiones, which in their properties differ from the corresponding hydroiodides. N-substituted aminoindones decomposed under the action of HCl. The imine hydrochlorides are dark-red or violet substances that form the starting imines with water. In the IR spectrum (3 and 6 μ regions) the presence of bands at 1597 cm\(^{-1}\), 1620 cm\(^{-1}\) (for III, \(R_1 = H\), \(R_2 = CH_3\), also at 1718 cm\(^{-1}\), probably due to an admixture of form II or traces of methylindandione) and 3150 cm\(^{-1}\) is characteristic (see the experimental part).

It is clear that the molecule contains neither a \(\mathrm{C=O}\) group nor an unconjugated \(\mathrm{C=N<}^{\oplus}\) group. The band at 3150 cm\(^{-1}\) may be assigned to associated stretching vibrations of the \(-\mathrm{OH}\) group, since hydroiodides of structure II have no such bands above 3000 cm\(^{-1}\). Similar absorption is also characteristic of hydroiodides of 1-aminocyclohexen-1-ones-3 \(^{(9)}\). Accordingly, the true structure of the hydrochlorides should be sought between the two limiting formulas IIIa and IIIb, \(R_1 = H\). Owing to delocalization of the positive charge, exact assignment of the bands in the 6 μ region to vibrations of definite bonds is hardly possible.

It seems to us that the chlorides and hydroiodides differ in their structure for crystallochemical reasons. First, methyl or phenyl substituents in position 2 of the five-membered ring and an alkyl or aryl substituent at the nitrogen atom, being different in their electron-donating properties, do not affect the structure of the substances. Second, the hydroiodide of the imine of methylindandione occupies, as it were, an intermediate position between the two groups of hydrohalides. Under certain conditions this compound can be obtained in a red form, which in solutions, in the presence of a large excess of HJ, spontaneously changes into the yellow form. The yellow form of the hydroiodide always has a clearly expressed additional band at 1629 cm\(^{-1}\), which must be assigned to the red form. The red form of this hydroiodide is very unstable, and it has not been possible to obtain it in pure form.

• Solution in dichloroethane.

Table 2

Infrared spectra of the hydroiodides

\[ \begin{array}{c} \overset{\oplus}{\mathrm{NHR_1}}{=}C \left(-C_6H_4-\mathrm{CO}-\right) \left(-C(R_2)(R_3)-\right)J^{\ominus} \end{array} \]

spectra taken on an IKS-14 instrument; vaseline oil

it proved possible. At the same time, all the other hydroiodides exist only in the yellow form, whereas the hydrochlorides obtained by us are in the red form.

The formation of free aminoindones (I) from their salts in the presence of basic solvents (alcohol, water) is understandable. Salts of aminoindones, depending on their structure, behave either as strong N- or O-acids (III), like the salts of amides ($^{17}$), or as strong C-acids. In our case the proton in the hydroiodides (II) is strongly activated by the electrophilic carbonyl and immonium groups.

As for the phenomenon in which 2-phenylindandione imine in conc. H$_2$SO$_4$ forms a colorless or yellowish solution ($^{18}$), here, evidently, non-isolable salts of the imine with two equivalents of acid are formed (cf. ($^{10}$)), for example, of type (VI).

Experimental Part

Hydroiodides of aminoindones (IIa–zh)

Hydrogen iodide is passed, with cooling, through a suspension of 0.25 g of aminoindone (Ia–zh) in 5 ml of CH$_3$COOH. The aminoindone gradually dissolves, and crystals of the hydroiodide separate from the dark-brown (sometimes dark-orange) solution. The salt is filtered on a Hirsch funnel, washed with several drops of CH$_3$COOH, and dried (see Table 1). Washing the salts with ether or crystallization from CH$_3$COOH is permissible only in the case of salts of 3-alkylaminoindones. Sometimes isolation of the substance requires very rapid operations because of the sensitivity of the product to atmospheric moisture. For example, the upper layer of the freshly filtered substance has to be removed. In such cases the yield is not given in Table 1 and below.

In the case of salt IIzh, 6 ml of a mixture of CH$_3$COOH and CH$_2$ClCH$_2$Cl (1:1), or only CH$_2$ClCH$_2$Cl, are taken, and HJ is passed in until the mixture becomes light yellow. The salts are not...

slightly soluble in \(CH_2ClCH_2Cl\), \(CH_3COOH\), and some—in acetone.

Hydrochloride of methylindandione imine (III, \(R_1 = H;\ R_2 = CH_3\)). HCl is passed through a cooled solution of 0.25 g of imine I (\(R_1 = H;\ R_2 = CH_3\)) in 5 ml of THF (or \(CH_3COOH\)) until precipitation of the hydrochloride ceases. The salt is filtered off, washed with THF, and dried over \(P_2O_5\). Dark-red crystals are obtained which, with water, form an orange-red solution; from the solution, the methylindandione imine is almost immediately precipitated. IR spectrum, in cm\(^{-1}\): 1585 (88), 1620(81), 1718(66), 3155(79).

Found, %: N 7.37; Cl 17.89
\(C_{10}H_9ON \cdot HCl\). Calculated, %: N 7.16; Cl 18.07

Hydrochloride of phenylindandione imine (III, \(R_1 = H;\ R_2 = C_6H_5\)). A suspension of 0.1 g of imine I (\(R_1 = H;\ R_2 = C_6H_5\)) in 5 ml of \(CH_3COOH\) is saturated with dry HCl under cooling. To the dark-brown solution, 20 ml of abs. ether is gradually added, rubbing the walls of the flask with a rod. From the orange solution a violet salt precipitates, which is filtered off as rapidly as possible on a Hirsch funnel and washed with several drops of dry ether. It is dried over \(P_2O_5\) in the presence of \(PCl_5\). IR spectrum in cm\(^{-1}\): 1585(63); 1620(51); 3157(63).

Found, %: N 5.23; Cl 13.85
\(C_{15}H_{11}ON \cdot HCl\). Calculated, %: N 5.44; Cl 13.76

Preparation of the hydroiodide (IIж). a) 0.25 g of imine I (\(R_1 = H;\ R_2 = CH_3\)) is dissolved on heating in 6 ml of a mixture of \(CH_3COOH\) and \(CH_2ClCH_2Cl\) (1 : 1). HI is passed through the solution. A dark-red substance precipitates, which then passes into the yellow form. b) A strong stream of HI is passed for a short time through a saturated solution of I (\(R_1 = H;\ R_2 = CH_3\)) in dichloroethane. Dark-red, lustrous crystals of the hydroiodide separate; they are dried over \(P_2O_5\). The red form of the hydroiodide, like the yellow form, is soluble in water with an orange-red coloration; from the solution, imine (I) is rapidly precipitated. The substance contains approximately 1 molecule of HI per 1 molecule of methylindandione imine. Absorption of one of the samples in cm\(^{-1}\): 1597(73); 1626(69); 1659(72), 1720(61, shoulder).

The starting aminoindones were prepared by the following methods (I, Table 1): a, б, в (\(^{19}\)), г (\(^{11}\)), д (\(^{20}\)), ж (\(^{18}\)).

Institute of Organic Synthesis
Academy of Sciences of the Latvian SSR

Received
14 X 1961

CITED LITERATURE

1. W. M. Lauer, G. W. Lones, J. Am. Chem. Soc., 59, 232 (1937).
2. A. Combes, Bull. Soc. Chim., 7 (3), 778 (1892).
3. R. Robinson, J. Chem. Soc., 109, 1038 (1916).
4. Н. К. Кочетков, Изв. АН СССР, ОХН, 1954, 47.
5. G. Eglington, M. C. Whiting, J. Chem. Soc., 1953, 3052.
6. R. Wizniger, U. Arni, Chem. Ber., 92, 2309 (1959).
7. Z. Arnold, F. Šorm, Coll. Czechoslov. Chem. Commun., 23, 492 (1958).
8. G. Hager, J. S. Hanker, J. Am. Pharmac. Assoc., Sci. Ed., 44, 138 (1955); Zbl., 1958, 2111.
9. N. J. Leonard, J. A. Adamcik, J. Am. Chem. Soc., 81, 595 (1959).
10. T. G. Bonner, M. Barnard, J. Chem. Soc., 1958, 4176.
11. Я. Ф. Фрейманис, Г. Я. Ванаг, ЖОХ, 30, 3369 (1960).
12. D. G. O’Sullivan, J. Chem. Soc., 1960, 3278.
13. C. H. de Puy, P. R. Wells, J. Am. Chem. Soc., 82, 2903 (1960).
14. А. К. Арен, Г. Я. Ванаг, ЖОХ, 31, 117 (1961).
15. A. Mustafa, A. H. E. Harhash, J. Am. Chem. Soc., 78, 1649 (1956).
16. B. Witkop, J. Am. Chem. Soc., 78, 2873 (1956).
17. А. Е. Куликова, Е. Н. Зильберман, Н. А. Сазонова, ЖОХ, 30, 2180 (1960).
18. Я. Ф. Фрейманис, Г. Я. Ванаг, ЖОХ, 30, 3362 (1960).
19. G. Wanag, U. Walbe, Ber., 69, 1054 (1936).
20. Г. Я. Ванаг, Я. Ф. Фрейманис, Г. Ф. Закис, ЖСХ, 27, 2509 (1957).