B. V. Tronov, L. P. Kulev, and V. N. Orlova

Reactivity of Halogens in Halogen-Substituted Ketones

Halogen derivatives of acetone, methylethyl ketone, pinacolin, acetophenone, cyclohexanone, diisobutyl ketone, and methylacetylacetone were taken, containing chlorine or bromine in primary, secondary, or tertiary units. For comparison, experiments were carried out to determine the reactivity (under the same conditions) with the ethyl esters of chloroacetic and bromoacetic acids.

Some halogen-substituted compounds were taken ready-made (Kahlbaum or Schuchardt preparations, which were purified before investigation), while the greater part were prepared by halogenation of the ketones listed above.

Of the preparative part of the work, we shall present here only the experiments on bromination of diisobutyl ketone, methylacetylacetone, and acetylacetone, since with these ketones substances not yet described in the literature were obtained.

Bromination of diisobutyl ketone. To 17 g of the ketone, 17 g of bromine was added under cooling with ice water. The hydrogen bromide evolved was removed by passing CO₂ through for 7 hours. The remaining liquid (24 g) was distilled at 75–76 mm. The following fractions were collected:

Analysis of the last fraction was carried out by heating the substance with an excess of sodium ethylate in alcoholic solution on a water bath. The separated ionized halide was determined by the gravimetric method. To ensure complete decomposition, one sample was heated for 8 hours, another for 16 hours. Results of the analysis:

Sample of 0.1781 g gave 0.2172 g AgBr.
=0.1570 g gave 0.1964 g AgBr.

For C₉H₁₆Br₂O, Br required—53.3%.
Found: I—52.53%; II—53.24%.
Fraction III gave 45.40% bromine.

Thus, the product obtained proved to be a dibromo-substituted ketone. As regards the position of the bromine atoms, according to literature data—

...one may think that they are either both at one carbon atom adjacent to the carbonyl group, or on different sides of the carbonyl:

\[ \begin{array}{ccccc} & \mathrm{CH_3} & & & \mathrm{CH_3}\\ & \diagdown & & & \diagup\\ \mathrm{CH_3} & & \mathrm{CH}\cdot\mathrm{CBr_2}\cdot\mathrm{CO}\cdot\mathrm{CH_2}\cdot\mathrm{CH} & & \mathrm{CH_3} \end{array} \]

or

\[ \begin{array}{ccccc} & \mathrm{CH_3} & & & \mathrm{CH_3}\\ & \diagdown & & & \diagup\\ \mathrm{CH_3} & & \mathrm{CH}\cdot\mathrm{CHBr}\cdot\mathrm{CO}\cdot\mathrm{CHBr}\cdot\mathrm{CH} & & \mathrm{CH_3} \end{array} \]

By the action of bromine on the first fraction of the distillate, a liquid of composition \(\mathrm{C_9H_{16}Br_2O}\) was likewise obtained (analysis gave 52.76% bromine), but it differed noticeably from the first dibromo-diisobutyl ketone in specific gravity.

\[ \text{I) } d^{20}_{4}=1.42 \qquad\qquad \text{II) } d^{20}_{4}=1.74. \]

The activity of the haloid in these compounds also proved to be very different.

To the second product of bromination, judging from the method of its preparation, one may assign the structure

\[ \begin{array}{ccccc} & \mathrm{CH_3} & & & \mathrm{CH_3}\\ & \diagdown & & & \diagup\\ \mathrm{CH_3} & & \mathrm{CBr}\cdot\mathrm{CHBr}\cdot\mathrm{CO}\cdot\mathrm{CH_2}\cdot\mathrm{CH} & & \mathrm{CH_3} \end{array} \]

(if one admits that at first a monosubstituted compound was formed, which on distillation split off HBr, and on further bromination added two atoms of halogen).

Nor is one of the preceding structures excluded (if the first fraction of the distilled liquid was unchanged diisobutyl ketone, which then for some reason was brominated in another direction).

The bromination of methylacetylacetone was carried out in the presence of water and calcium carbonate at room temperature. The oil that separated was washed with a solution of soda until decolorized, then dried with calcium chloride. The substance resinifies very readily and could not be distilled at all.

The product purified as far as possible was a yellowish liquid with a sharp odor. Specific gravity

\[ d^{18}_{4}=1.58. \]

Determination of the haloid gave the following results:

Sample \(0.0731\) g; AgBr—\(0.0744\) g.
\(0.3943\) g; » —\(0.3946\) g.

For \(\mathrm{C_6H_9BrO_2}\) Br—41.45% is required; 41.82% and 42.05% were found.

According to all the literature data and the activity of bromine, the most probable structure is

\[ \mathrm{CH_3\cdot CO\cdot CBr\cdot CO\cdot CH_3} \]
\[ \phantom{\mathrm{CH_3\cdot CO\cdot}}\big| \\ \phantom{\mathrm{CH_3\cdot CO\cdot}}\mathrm{CH_3} \]

We also carried out the bromination of acetylacetone. Addition of bromine to a solution of the ketone in petroleum ether gave two products: a solid, insoluble in petroleum ether, and a liquid, more readily soluble. For the first, analysis showed the following halide content:

Sample 0.9757 g; AgBr—1.0425 g.
» 0.3657 » —0.3779
» 0.1472 » —0.1521

For \(C_5H_7O_2Br\), Br required—44.69%.
Found: 45.44%; 44.00%; 44.11%.

Analysis of the liquid substance:

Sample 0.7624; AgBr—1.0959 g.

For \(C_5H_6O_2Br_2\), 62.02% Br required; found 61.11%.

Thus, the composition of the compounds obtained corresponds to monobromoacetylacetone and dibromoacetylacetone. In view of the extreme instability of both products, no tests of their activity were carried out with them.

Determination of Halide Activity

Since the halide in almost all the ketones we studied is split off extremely readily, of the reagents ordinarily used it was necessary in almost all cases to take only the weaker one, pyridine. The measurement procedure was also changed: the reaction was almost always conducted in benzene solution. Equimolecular amounts of ketone and amine were taken and were first dissolved in benzene, calculated so that after mixing a concentration of 1 gram-molecule per liter would be obtained. The temperature of the experiments was 17–19°. The work-up was the usual one, i.e., the mixture was diluted with water, the unreacted ketone was extracted with organic solvents, and the ionized halide was precipitated with silver nitrate.

The accuracy of the determination was impaired by the fact that some halogen-substituted ketones dissolve in water, which makes their removal difficult. Symmetrical dichloroacetone proved even poorly soluble in benzene; experiments with it had to be carried out in methyl alcohol solution. For comparison, this was also done with chloroacetone.

The results of the determination are expressed in the following tables:

Experiments with pyridine in benzene solution:

Product—oil and crystals

1. \(CH_3.CO.CH_2Cl\).

Product crystall.

1. $\mathrm{CH_3.CO.CH_2Br}$

Product—oil

1. $\mathrm{CH_3.CO.CHCl_2}$.

Product crystall.

1. $\mathrm{CH_3.CH_2.CO.CH_2Cl}$.

Product oil

1. $\mathrm{CH_3.CO.CHCl.CH_3}$.

Product crystall.

1. $\mathrm{CH_3.CH_2.CO.CH_2Br}$.

Product oil

1. $\mathrm{CH_3.CO.CHBr.CH_3}$.

Product crystall.

1. $(\mathrm{CH_3})_3\mathrm{C.CO.CH_2Br}$.

Product crystalline.

9. $C_6H_5 \cdot CO \cdot CH_2Cl.$

Product crystalline.

10. $C_6H_5 \cdot CO \cdot CH_2Br.$

11.

\[ \begin{array}{c} \mathrm{CH_2{-}CHCl}\\[-2mm] \mathrm{CH_2}\!<\!{\phantom{xx}}\!>\!\mathrm{CO}\\[-2mm] \mathrm{CH_2{-}CH_2} \end{array} \]

12.

\[ \begin{array}{c} \mathrm{CH_2{-}CHBr}\\[-2mm] \mathrm{CH_2}\!<\!{\phantom{xx}}\!>\!\mathrm{CO}\\[-2mm] \mathrm{CH_2{-}CH_2} \end{array} \]

13. $CH_2Cl \cdot COOC_2H_5.$

Product crystalline.

14. $CH_2Br \cdot COOC_2H_5.$

Experiments with pyridine in methyl alcohol

15. $CH_3 \cdot CO \cdot CH_2Cl.$

16. $CH_2Cl \cdot CO \cdot CH_2Cl.$

Experiments with pyridine without solvent

17. Dibromo-diisobutyl ketone I.

18. Dibromo-diisobutyl ketone II.

19. $\mathrm{CH_3 \cdot CO \cdot CBr(CH_3) \cdot CO \cdot CH_3}$.

20. Experiments with $\mathrm{CH_3ONa}$ in methyl alcohol.

Dibromo-diisobutyl ketone I. (molar ratio 1:1).

21. Dibromo-diisobutyl ketone I (1 mol ket. + 2 mol $\mathrm{CH_3ONa}$).

22. $\mathrm{CH_3 \cdot CO \cdot CBr(CH_3) \cdot CO \cdot CH_3}$.

In the following summary table, the numbers of the relative reaction rates are compared. The constants were not calculated; the comparison was made by the intervals of time during which the same percentage of halide reacts. The reaction rate of methylchloroethyl ketone with pyridine in benzene solution was taken as unity. Average values were calculated for the intervals 0–5%, 5–10%, 10–20%, 20–30%, 30–40%, 40–50%.

Only data from experiments with pyridine are included in the table.

*) Calculated for all the halide contained in the molecule.

Table 23

Relative reaction rates

$^{1)}$ Taken from the article by B. V. Tronov, A. I. Akivis, and V. N. Orlova, Zh. R. Kh. O. 61, 345–353.

CONCLUSIONS:

1) Table 24

Influence of the nature of the halide

Bromine is everywhere much more active than chlorine. The difference is comparatively small in the secondary halogen derivatives of methyl ethyl ketone.

Influence of the accumulation of halogen atoms

The ratio of the rates of reaction with pyridine for $\mathrm{CH_3CO\cdot CHCl_2}$ and $\mathrm{CH_3\cdot CO\cdot CH_2Cl}$ is on average 0.4. Thus, the second halogen atom in the same linkage decreases the activity. The second chlorine in position 1,3 (and here also on the other side of the carbonyl) makes the substance, on the contrary, so active that any exact measurement becomes impossible. The ratio of the mobility of the halide in $\mathrm{CH_2Cl\cdot CO\cdot CH_2Cl}$ and $\mathrm{CH_3\cdot CO\cdot CH_2Cl}$ may be estimated approximately by a number of the order of $10^3$.

3) Influence of the position of the halide in a primary, secondary, or tertiary linkage

On this question, in the present work there are only very few data.

Table 25

With the same structure of the molecule, the halide in the primary linkage is more active with respect to amines.

4. Effect of the Size of the Radical

Table 26

With one and the same group $\mathrm{CH_2X \cdot CO}$, as the other radical of the ketone is increased, the activity in the greater part of the aliphatic-series compounds taken by us decreases. Halogenated acetophenones react almost as readily as haloacetones¹.

5) Effect of Ring Closure

Here an exact comparison is as yet impossible, since ketones of somewhat different chain length have to be taken: with 5 and 6 carbons.

Table 27

Ring closure causes a sharp fall in the activity of the halide (with the same position in the secondary chain, adjacent to the carbonyl group) relative to pyridine.

6. Comparison of the Influence of the Ketone and Ester Groups

Table 28

Analogously to what was noted earlier¹ in the example of halogen-substituted esters of ketonic acids and dicarboxylic acids, the ketone group $\mathrm{CH_3 \cdot CO}$ imparts greater mobility to the halogen than does the group $\mathrm{C_2H_5O \cdot CO}$.

7) Influence of the Solvent

Comparison of the results of experiments with bromoacetoacetic ester without solvent and in benzene medium shows that in a solution of concentration 1N the reaction proceeds thousands of times more slowly. In methyl alcohol, chloroacetone and pyridine react 4–7 times faster than in benzene.

Type of Cleavage of the Bond between Halogen and Carbon

For determining the electronic character of the activity of halogens in halogen-substituted ketones, our experiments provide very little material, since the reaction rate was almost everywhere determined only with pyridine. An experiment with $\mathrm{CH_3ONa}$ gave, for brominated acetylacetone, an activity exceeding the activity of this same ketone toward pyridine (without solvent) in any case by several thousand times, i.e., the same as is observed for tertiary halogen derivatives of hydrocarbons² and esters. Obviously, the distribution of compounds into activity groups is everywhere the same: primary halogen-substituted compounds react more readily with amines and halide salts³, tertiary ones—with alkalis and alcoholates.

17/VII—1930.

Tomsk, Laboratory of Organic Chemistry of the Siberian Technological Institute and Chemical Laboratory of the 1st Siberian Pharmaceutical Technical School.

¹ B. V. Tronov, A. I. Akivis and V. N. Orlova.
² B. V. Tronov and L. G. Boldyreva. Work in press.
³ For experiments with KI see J. B. Conant, R. E. Hussey, J. Am. Ch. Soc. 47, 476–488 (1925).
J. B. Conant, W. R. Kirner, R. E. Hussey. J. Am. Ch. S. 47, 488–501.

1. It must also be taken into account that, owing to solubility conditions, experiments with chloro- and bromoacetones may have given a somewhat exaggerated result. Part of the halogen-substituted compound could remain in the aqueous solution and then react with $\mathrm{AgNO_3}$. ↩