Full Text
Chemistry
M. A. KHAIMOVA and B. I. KURTEV
ON THE INTERACTION OF CERTAIN ARYLATED ETHYLENIC HYDROCARBONS WITH IODINE CHLORIDE AND ON THE PINACOLINE-TYPE REARRANGEMENT OF THEIR 1-IODO-2-METHOXY DERIVATIVES
(Presented by Academician B. A. Kazanskii on 30 VI 1960)
As Temnikova, Baskova, and Khaimova showed \(\left(^{1}\right)\), iodine chloride reacts differently with 1,1-diphenylethene (I) and with 1,1-diphenyl-1-propene (II): in the first case normal addition according to Markovnikov’s rule takes place, whereas from hydrocarbon II, instead of the expected 1-chloro-2-iodo-1,1-diphenylpropane, 2-chloro-1,1-diphenyl-1-propene was obtained. The reaction was accompanied from the very beginning by the liberation of iodine and hydrogen chloride, and for its completion two moles of iodine chloride were required. It was initially assumed that II and iodine chloride form 1-iodo-2-chloro-1,1-diphenylpropane, which spontaneously eliminates hydrogen iodide, which with iodine chloride gives iodine and hydrogen chloride.
Addition not in accordance with Markovnikov’s rule was ascribed to the influence of the methyl group. However, an investigation of the interaction of iodine chloride with 1,1-di-(\(p\)-tolyl)-1-propene (III), carried out by one of us at Temnikova’s suggestion \(\left(^{2}\right)\), showed that the latter hydrocarbon behaves analogously to II. Since it is difficult to suppose that the orienting influence of the methyl group on the course of addition of iodine chloride can predominate over the influence of two phenyl and, in particular, two \(p\)-tolyl groups, we began to seek another explanation of the final results. The only proof for the addition of iodine chloride to hydrocarbons II and III not in accordance with Markovnikov’s rule was the position of the chlorine atom in the isolated unsaturated monochlorides. In the literature, however, examples have been described of replacement of an iodine atom by a chlorine atom under the action of iodine chloride \(\left(^{3,4}\right)\). If such replacement also occurs in our cases, then this proof proves invalid. One of us \(\left(^{5}\right)\) showed that in the interaction of I with two moles of iodine chloride, after the addition of the first mole of iodine chloride, liberation of iodine and hydrogen chloride begins, and as a result the product obtained is no longer 1-chloro-2-iodo-1,1-diphenylethane, but 2-chloro-1,1-diphenylethene.
We studied the interaction of iodine chloride also with trans-1,2-diphenyl-1-propene (\(\alpha\)-methylstilbene, IV). In this case, even when cooling was applied from the very start of the reaction, liberation of iodine was observed. Hydrogen chloride, however, was liberated only toward the end of the reaction. From the reaction mixture, 1,2-dichloro-1,2-diphenylpropane was isolated (probably in the form of a mixture of diastereomers), which, on treatment with alcoholic alkali, gave two crystalline, already described isomers of 1-chloro-1,2-diphenyl-1-propene (the low-melting isomer—only in the form of an oil) \(\left(^{6}\right)\). Consequently, the liberation of iodine and the liberation of hydrogen chloride are the results of two different reactions.
The facts set forth above led us to adopt the following reaction scheme, general for the interaction of all the hydrocarbons I—IV studied with iodine chloride:
\[ \begin{gathered} \begin{array}{cccccc} & \mathrm{Ar} & & \mathrm{Ar}\ \ \mathrm{Cl}\ \ \mathrm{J} & & \mathrm{Ar}\ \ \mathrm{Cl}\ \ \mathrm{Cl} \\ & \backslash & & \backslash\ \ | \ \ | & & \backslash\ \ | \ \ | \\ & \mathrm{C}=\mathrm{CH}-\mathrm{R}_2 & \xrightarrow{+\mathrm{JCl}} & \mathrm{C}-\mathrm{CH}-\mathrm{R}_2 & \xrightarrow[\mathrm{J}_2]{+\mathrm{JCl}} & \mathrm{C}-\mathrm{CH}-\mathrm{R}_2 \xrightarrow{-\mathrm{HCl}} \begin{array}{c} \mathrm{Ar}\ \ \mathrm{Cl}\\ \backslash\ \ |\\ \mathrm{C}=\mathrm{C}-\mathrm{R}_2 \end{array} \\ & / & & / & & / \\ & \mathrm{R}_1 & & \mathrm{R}_1 & & \mathrm{R}_1 \end{array} \qquad (1) \end{gathered} \]
\[ \begin{array}{c|ccc|ccc} & \mathrm{Ar} & \mathrm{R}_1 & \mathrm{R}_2 & & \mathrm{Ar} & \mathrm{R}_1 & \mathrm{R}_2\\ \mathrm{I} & \mathrm{C}_6\mathrm{H}_5 & \mathrm{C}_6\mathrm{H}_5 & \mathrm{H} & \mathrm{III} & n\text{-}\mathrm{CH}_3\text{-}\mathrm{C}_6\mathrm{H}_4 & n\text{-}\mathrm{CH}_3\text{-}\mathrm{C}_6\mathrm{H}_4 & \mathrm{CH}_3\\ \mathrm{II} & \mathrm{C}_6\mathrm{H}_5 & \mathrm{C}_6\mathrm{H}_5 & \mathrm{CH}_3 & \mathrm{IV} & \mathrm{C}_6\mathrm{H}_5 & \mathrm{CH}_3 & \mathrm{C}_6\mathrm{H}_5 \end{array} \]
From what has been presented it is clear that the relative rate of the individual stages of the process and the stability of the intermediate products depend strongly on the structure of the starting hydrocarbon. With hydrocarbons II—IV the slowest stage is the first, and therefore in these cases it is not possible to isolate the 1-chloro-2-iodo derivatives, which would make it possible to judge the order of addition of iodine chloride.
To confirm scheme (1), it was necessary to establish the order of addition of other polar molecules to II, since in the literature there was not sufficiently reliable information for solving this question (⁷). Following Tiffeneau’s example (⁸), we tried to convert II into the iodohydrin and to treat the latter with silver nitrite. Depending on the structure of the iodohydrin, the following reactions could be expected:
\[ \begin{gathered} \begin{array}{c} \mathrm{C}_6\mathrm{H}_5\\ \backslash\\ \mathrm{C}=\mathrm{CH}-\mathrm{CH}_3\\ /\\ \mathrm{C}_6\mathrm{H}_5 \end{array} \ \begin{array}{c} \nearrow\\[-2mm] \searrow \end{array} \left[ \begin{array}{ccc} \mathrm{C}_6\mathrm{H}_5 & \mathrm{OH}\cdot\mathrm{J} & \\ \backslash & | & |\\ & \mathrm{C}-\mathrm{CH}-\mathrm{CH}_3 & \\ / & & \\ \mathrm{C}_6\mathrm{H}_5 & & \end{array} \right] \xrightarrow{-\mathrm{HJ}} \mathrm{C}_6\mathrm{H}_5-\mathrm{CO}-\mathrm{CH}(\mathrm{CH}_3)-\mathrm{C}_6\mathrm{H}_5 \qquad (2a) \\[4mm] \left[ \begin{array}{ccc} \mathrm{C}_6\mathrm{H}_5 & \mathrm{J} & \mathrm{OH}\\ \backslash & | & |\\ & \mathrm{C}-\mathrm{CH}-\mathrm{CH}_3 & \\ / & & \\ \mathrm{C}_6\mathrm{H}_5 & & \end{array} \right] \xrightarrow{-\mathrm{HJ}} \begin{array}{c} \mathrm{C}_6\mathrm{H}_5\\ \backslash\\ \mathrm{CH}-\mathrm{CO}-\mathrm{CH}_3\\ /\\ \mathrm{C}_6\mathrm{H}_5 \end{array} \qquad (2b) \end{gathered} \]
A preliminary experiment, however, gave a carbonyl compound in an unsatisfactory yield. Taking into account that β-iodo ethers are more stable than the corresponding iodohydrins, we treated II with iodine in the presence of yellow mercuric oxide and methyl alcohol and then treated the product obtained with a solution of silver nitrite. From the reaction mixture we isolated only D,L-methyldesoxybenzoin in a yield of about 78% of theory. This shows that the intermediate β-iodo ether had the structure 1-methoxy-2-iodo-1,1-diphenylpropane, i.e., that the addition of the iodine atom and the methoxy group to II proceeded according to the classical Markovnikov rule (see scheme 2a).
In an analogous way, desoxybenzoin was obtained from I, i.e., the same result that had been obtained through its iodohydrin (⁸).
Only from I was it possible to obtain the crystalline β-iodo ether in pure form. The same product was obtained more conveniently from I, iodine chloride, and methanol.
The interaction of 1-methoxy-2-iodo-1,1-diphenylethane with an excess of iodine chloride in a methanolic medium led to its partial conversion into desoxybenzoin (in chloroform, the main reaction was replacement of the iodine atom by a chlorine atom, while the rearrangement was a side reaction). This experiment shows that iodine chloride is capable of causing pinacolone-type rearrangements by elimination of hydrogen iodide from compounds of this type. This indirectly explains the formation of unsaturated monochlorides with a rearranged carbon skeleton,
formed in small amounts in the reaction of II \(^{(1)}\) or III \(^{(2)}\)* with iodine chloride, as is seen from the following scheme:
\[ \left[ \begin{array}{c} \mathrm{Ar}\quad \mathrm{Cl}\quad \mathrm{J}\\ \quad \backslash\ \vert\ \vert\\ \quad \mathrm{C}-\mathrm{CH}-\mathrm{R}_2\\ \quad /\ \\ \mathrm{R}_1 \end{array} \right] \ \xrightarrow[\left(-\mathrm{HJ}\right)]{+\mathrm{JCl}}\ \mathrm{R}_1-\underset{\vert}{\overset{\mathrm{Cl}}{\mathrm{C}}}=\mathrm{C}-\mathrm{R}_2 \quad \underset{\mathrm{Ar}}{\vert} \tag{3} \]
Experimental Part
Reaction of 1,2-diphenyl-1-propene (IV) with iodine chloride. To a solution of 4.85 g (0.025 g-mol) of \(\alpha\)-methylstilbene (m.p. 81.5–82.5° \(^{(8)}\)) in 25 ml of dry chloroform, 8.13 g (0.050 g-mol) of iodine chloride solution is added dropwise. Hydrogen chloride gradually begins to be evolved. On the following day the reaction mixture is washed with sodium thiosulfate solution. After the chloroform is distilled off in vacuo, 6.55 g of a mixture of crystals and oil remains. The oil is washed away with cold methanol, and the crystals are recrystallized repeatedly from alcohol. A colorless crystalline substance with m.p. 104–106° is obtained.
\[ \begin{array}{r} \text{Found, \%: } \mathrm{C}\ 67.30;\ 67.50;\ \mathrm{H}\ 5.42;\ 5.32;\ \mathrm{Cl}\ 27.09;\ 27.00\\ \mathrm{C}_{15}\mathrm{H}_{14}\mathrm{Cl}_2.\ \text{Calculated, \%: } \mathrm{C}\ 67.93;\ \mathrm{H}\ 5.32;\ \mathrm{Cl}\ 26.74 \end{array} \]
A portion of the crystals (1,2-dichloro-1,2-diphenylpropane) is heated for 2 hours on a boiling water bath with a 10% alcoholic solution of caustic potash. From the residue after distilling off the alcohol, a mixture of crystals and oil is extracted with benzene. To increase the amount of crystals, the mixture is heated according to \(^{(8)}\) at 120° for 15 hours. From the reaction mixture two colorless products are isolated, purified by recrystallization from glacial acetic acid. The first of them melts at 121–122.5° and shows no depression of the melting temperature in admixture with 1-chloro-1,2-diphenyl-1-propene, described in \(^{(6)}\). The second product has m.p. 50–51° and is an isomer of the first.
\[ \begin{array}{r} \text{Found, \%: } \mathrm{Cl}\ 15.40;\ 15.15\\ \mathrm{C}_{15}\mathrm{H}_{13}\mathrm{Cl}.\ \text{Calculated, \%: } \mathrm{Cl}\ 15.50 \end{array} \]
The analyzed intermediate crystalline or liquid fractions showed the same chlorine content.
The low-melting 1-chloro-1,2-diphenyl-1-propene consists of long needles, readily soluble in ordinary organic solvents, somewhat less soluble in alcohol and in glacial acetic acid. It can also be obtained according to \(^{(6)}\). From the oily reaction product the same unsaturated monochlorides were isolated in an analogous manner.
Conversion of 1,1-diphenyl-1-propene (II) into 1,2-diphenylpropanone (I). To a mixture of 1.94 g (0.01 g-mol) of II (m.p. 48–49.5°), 0.012 g-mol of freshly prepared dry yellow mercuric oxide, and 25 ml of methanol, 0.02 g-mol of iodine dissolved in 30 ml of methanol is added dropwise with stirring. The mixture is stirred for 3 hours. On the following day the precipitate is filtered off, the filtrate is concentrated by evaporating part of the methanol, and concentrated aqueous silver nitrate solution is added to the residue, whereupon silver iodide immediately precipitates. About 1.65 g (78% of theory) of ketone with m.p. 48–50° separates from the solution. A mixed sample with an authentic specimen of \(D,L\)-methyldesoxybenzoin \(^{(10)}\) shows no depression of the melting temperature. The ketone was also identified in the form of semicarbazones according to \(^{(11)}\).
Synthesis of 1-methoxy-2-iodo-1,1-diphenylethane from 1,1-diphenylethene (I). a) By the method described, the reaction is carried out
* The experiment was repeated by us, the products being isolated by chromatography over alumina, and not by distillation.
with iodine and methanol in the presence of yellow mercuric oxide 4.14 g of hydrocarbon I. The mixture is stirred for several hours, then diluted with water, the precipitate of mercuric iodide is removed by filtration, and the filtrate is extracted with chloroform. The residue after removal of the chloroform is recrystallized from alcohol. This gives 5.94 g (76% of theory) of 1-methoxy-2-iodo-1,1-diphenylethane, m.p. 86–87°. The product recrystallized for analysis melted at 86–87.5°.
\[ \begin{array}{rll} \text{Found, \%:} & \mathrm{C}\ 53.54;\ 53.46; & \mathrm{H}\ 4.63;\ 4.55 \\ \mathrm{C}_{15}\mathrm{H}_{15}\mathrm{OI}. \ \text{Calculated, \%:} & \mathrm{C}\ 53.27; & \mathrm{H}\ 4.47 \end{array} \]
1-Methoxy-2-iodo-1,1-diphenylethane consists of colorless long prisms, readily soluble in ordinary organic solvents. On storage for more than one week it begins to turn yellow and decompose.
b) To a suspension of 2.70 g (0.015 g-mole) of I in 20 ml of absolute methanol, a solution of 2.68 g (0.0165 g-mole) of iodine chloride in 30 ml of methanol is added dropwise with stirring. The mixture is stirred for a further 2 hours. The precipitated solid is filtered off and washed with methanol. Yield 2.05 g. M.p. 85–87°. A mixed melting-point test with 1-methoxy-1-iodo-1,1-diphenylethane from a) shows no depression of the melting point. From the methanolic filtrate a further 1.72 g of the same substance is isolated (total yield 74% of theory).
Rearrangement of 1-methoxy-2-iodo-1,1-diphenylethane. a) With silver nitrate: a solution of 1.01 g (0.003 g-mole) of 1-methoxy-2-iododiphenylethane in 15 ml of methanol is mixed with an aqueous solution of 0.68 g (0.004 g-mole) of silver nitrate. Silver iodide precipitates immediately. On the following day the precipitate is filtered off, the filtrate is diluted with water and extracted with chloroform. The residue after removal of the chloroform is recrystallized from methanol. This gives 0.40 g (68% of theory) of deoxybenzoin, m.p. 54–56° (identification was carried out by a mixed melting-point test and by conversion into the semicarbazone).
b) With iodine chloride: a mixture of 5.07 g (0.015 g-mole) of 1-methoxy-2-iodo-1,1-diphenylethane and 3.09 g (0.019 g-mole) of iodine chloride in 70 ml of methanol is stirred for 10 hours and left to stand for about another day and a half. The precipitate, 2.34 g, is unchanged starting material. Dilution of the filtrate with water and recrystallization of the precipitate from ethanol gives a further 1.96 g of starting material—85% in all.
From the mother liquors, by means of the dinitrophenylhydrazine method (12), it is possible to isolate 0.63 g (11% of theory) of the 2,4-dinitrophenylhydrazone of deoxybenzoin, m.p. 198–200°. The melting point of a mixture with an authentic sample shows no depression.
A control experiment carried out in the absence of iodine chloride gave 99% of the starting 1-methoxy-2-iodo-1,1-diphenylethane and no traces of the 2,4-dinitrophenylhydrazone of deoxybenzoin.
We express our gratitude to Prof. T. I. Temnikova for advice during the conduct of this investigation.
Institute of Organic Chemistry
Bulgarian Academy of Sciences
Received
20 V 1960
REFERENCES CITED
- T. I. Temnikova, Z. A. Baskova, M. A. Khaimova, ZhOKh, Coll. II, 874 (1953).
- M. A. Khaimova, ZhOKh, 25, 387 (1955).
- C. Friedel, R. Silva, Bull. Soc. Chim., 17, 537 (1872); Zbl., 1872, 548.
- H. Baganz, Ber., 87, 1373 (1954).
- M. A. Khaimova, Dokl. Bolg. AN, 10, 371 (1957).
- J. Sudborough, Ber., 25, 2237 (1892); J. Chem. Soc., 71, 224 (1897).
- R. Perrot, C. R., 202, 494 (1936).
- M. Tiffeneau, Ann. (8), 10, 351 (1907).
- H. Ley, Ber., 50, 249 (1917).
- V. Meyer, L. Oelkers, Ber., 21, 1297 (1888).
- A. McKenzie, R. Roger, J. Chem. Soc., 125, 844 (1923).
- S. Winstein, L. L. Ingraham, J. Am. Chem. Soc., 77, 1738 (1955).