The Conformational Preference of the Cyano Group1 - ACS Publications

4m ,JXKSEX. VOL. 27. 4-(p-Chlorophenyl)-4-methylcyclohexanone (II).-To a stirred and refluxing solution prepared from 0.76 g. of potassium metal and ...
0 downloads 0 Views 370KB Size
4606

RICKRORN . 4 m ,JXKSEX

4-(p-Chlorophenyl)-4-methylcyclohexanone (II).-To a stirred and refluxing solution prepared from 0.76 g. of potassium metal and 100 ml. of dry t-butyl alcohol was added 3.0 g. of diester X during 5 min. Refluxing was continued for 19 hr., then the reaction mixture was cooled to 30", 20 ml. of water was added, and the alcohol was removed maintaining a bath temperature of 60". The residue was hydrolyzed with 100 ml. of 33% sulfuric acid (by volume), and the product was isolated as described for the preparation of I. The crude product was dissolved in ether and the ether solution was washed twice with water before drying with magnesium sulfate. The ether was removed and distillation of the crude material at 0.8 mm., with a bath temperature 165" gave 1 g. of product which was crystallized from pure pentane a t -20". There was obtained after two crystallizations 0.67 g. of white needles, m.p. 37.9-38.9". A second crop of 0.05 g., m.p. 37.838.7" was obtained from the filtrates, total yield 37%. Anal. Calcd. for Cl3HI5C10: C, 70.10; H, 6.79. Found: C, 70.05; H, 6.94.

VOL. 27

Dipole Moments.-The dipole moments of the various ketones were run a t 25", in benzene solution, and the data are given in Table 11. The dipole moment apparatus has been described previously.19 The moments were calculated by essentially the method of Halverstadt and Kumler,20 utilizing an IBM 650 computer programmed as described earlier.21 The molar refractivities were obtained from tables,22 and atomic polarization wm neglected.

(19) N. L. Allinger, H. M. H a t t e r . M. A. DaRooge, and L. A. Freiberg, J . Org. Chem., 16, 2550 (1961). (20) I. F. Halverstadt a n d W. D. Kumler, J . A m . Chem. Soc., 64, 2988 (1942). ( 2 1 ) N. L. Allinger a n d J. Allinger, J . Org. Chem., S4, 1613 (1959). (22) A. I. Vogel, W. T. Cresswell, G. J. Jeffrey, a n d J. Leicester, Chem. Ind. (Lmdon), 358 (1950). (23) A recent value reported for this compound under identical conditions is 3.08 D; see H. H. Giinthard a n d T. Gaumann, Heh. chim. Acta, 34, 39 (1951).

The Conformational Preference of the Cyano Group1 BRUCERICK BORN^

APD

FREDERICK R. JENSEX

Department of Chemistry, University of California, Berkeley 4, California Received J u l y 16, 1968 The conformational preference of the cyano group has been determined by the base-catalyzed equilibration of 4-t-butylcyclohexanecarbonitrile. At 25' in t-butyl alcohol solvent, the equatorial cyano group is favored with AF = -0.15 kcal./ mole. The temperature dependence of the equilibrium indicates that this small free energy difference is due largely to a positive entropy term, A S = 0.7 =!= 0.3, ( A H = 0.04 f 0.1 kcal./mole) for the axial-equatorial interconversion.

Determinations of equatorialaxial conformational preferences in cyclohexane derivatives have provided valuable information regarding the nature of steric interactions in organic molecule^.^ The free energy difference (the Winstein A value)4 does not always parallel the covalent radius of the substituent. A low temperature n.m.r. study5 of the cyclohexyl halides, for example, has shown that the A values follow the order C1> Br > I > F. It was suggested5 that the conformational preference of a substituent depends on a combination of bond length, radius of the electron cloud, and the polarizability of this cloud. Such an explanation can account for the order observed with the halides, and also the small A value of the bromomercuri group.6 The question of electron density and polarizability also arises in substituted cyclohexanones. It has been assumed that the n-cloud of a carbonyl (1) (a) Supported in p a r t b y the National Science Foundation; (b) This work was originally submitted for publication elsewhere, a t which time it was learned t h a t Prof. N. L. Allinger had similar work accepted for publieation in J . Org. Chem. We wish t o express our appreciation t o Prof. Allinger for delaying publication of his work 60 t h a t t h e two papers could appear simultaneously. (2) Preeent address: University of California a t Santa Barbara, Goleta, Calif. (3) E. L. Eliel, "Stereochemistry of Carbon Compounds," McGrawHill Book Co., Inc.. F e w York, N. Y., 1962, p. 236. (4) S. Winstein a n d h-. J. Holness, J . A m . Chem. SOC.,77, 55fi2 (1955). (5) A. J. Berlin and F. R. Jensen, Chem. I n d . (London), 998 (1960). (6) F. R. Jensen a n d L. H. Gale, J . A m . Chem. Soc., 81, 6337 (1959).

group can be neglected in the calculation of steric interaction^,'^^ giving rise t o the "3-alkyl ketone effect." From the position of equilibrium for the geometrical isomers menthone and isomenthone, however, it has been estimated that the a-cloud of the carbonyl group is sterically comparable to an axial hydrogen. Important substituents for which the conformational preferences are unknown are those in which the group attached to the cyclohexane ring bears a triple bond. Included in this class are the sub-

+

+ -

stituents -C=N, -N=Y, -N=C , and -C=C-R. If, in these cases, the n-cloud is of appreciable size, rather large A-values for these groups are not unreasonable. For example, as an approximation, tJhe N2+ group has been regarded as sterically equivalent to the methyl group.'" Preliminary low temperature n.m.r. studies of cyclohexanecarbonitdel1 indicated a small conformational preference for the cyano group. Homever, the signals were not sufficiently separated to obtain an accurate value. An alternate method, using the bulky t-butyl group with its large equato( 7 ) W. Klyne, Eiperientia, 12, 119 (1956). (8) N. L. Allinger and L. A. Freiberg, J . Am. Chem. Soc.. 84, 2201 (1962). (9) B. Rickborn, i b i d . , 84, 2414 (1962). (10) A. Streitwieser, Jr., and W. D. Schaeffer, z h z d . . 79. 2888 (1957). T h e A-value for methyl is 1.97 kcal./mole, ref. 13. (11) Unpublished results of Mr. A . J, Berlin,

CONFORMATIONAL PREFEREKCE OF THE CYANO GROUP

DECEMBER, 1962 TABLE I

EQUILIBRATION OF Cis- AND trans-4-t-BuTYLCYc1,OHEXANECARBONITRI1.E WITH POTASSIUM &BUTOXIDE IN I-BUTYL

ALCOHOLSOLUTION T.

%

K

OC.

trans

trans/cis"

56.4 1.295 zk 0.017 56.5 1.301 zk 0.016 79.0b 56.7 1.309 zk 0.015 55.9 1.268 zkO.009 25.2' a Average deviations are reported. at room temperature with [t-BuO-] = = 0.3 Jf. 25.2b 4!).Hb

AF, kcal./mole"

-0.153 f 0.008 -0.169 f 0.009 -0.189 f 0.010 -0.141f0.005 Solutions prepared [t-RuO-] 1.0 M .

rial preference4 to fix the conformation of the cyano group, was chosen for the present study. The equilibrium constants for the interconversion of cis- and trans-4-t-butylcyclohexanecarbonitrile (equation 1) were determined a t three temperatures, and the values are shown in Table I.

In order to obtain the isomerically pure nitriles for this investigation, their preparation from the corresponding amides was studied using various dehydrating agents. Under the reaction conditions employed, both phosphorus pentoxide and thionyl chloride gave nitriles which, when analyzed by vapor phase chromatography, showed no detectable amount of the alternate isomer (>99% isomeric purity). With phosphorus oxychloride, however, some a-carbon isomerization was observed. The cis and trans nitriles used for the equilibration experiments were sharp-melting, isomerically pure compounds. Of particular interest is the observat'ion that the low A value for the cyano group (0.15 kcal./mole), is due mainly to the entropy term, AX = 0.66 f 0.3 e.u. As can be seen from the negligible effect of temperature on the equilbrium constant, the enthalpy term is small, and the calculated value 0.095 kcal./ is actually positive, AH = 0.042 mole. A zero entropy change for an axialequatorial interconversion has frequently been assiimed.12 Although conformational preferences I SI ally have been determined at only one temperature, d a h are available for cyclohexyl bromide in carbon disulfide solution a t 30" ( A F = -0.61 k c a l . / m ~ l e )and ~ ~ -81" ( A F = -0.48 kcal./ m ~ l e . ~ From , l ~ these values, A S E 1 e.u. and AH -0.3 kcal./mole for the axial-equatorial interconversion of the bromo group. From these (12) In some cases this assumption has been found to be valid, e . g . , for the dirnethylcyclohexanes: C. W. Beckett, K. 9. Pitzer, a n d R . Rriitaer. J . A m . Chem. SOC.. 6 9 , 2488 (1947). ( 1 3 ) F. R . Jensen and L. H. Gale, J. Org. Chem., 2S, 2075 (1960). (14) T h e vnluc selected for the n.1n.r. measurements a t -81' (rei. R ) which depends fin areits oi signals rather than chrniicul shifts is regarded as being the more relinble.

4607

results it can be seen that it is not necessarily valid to assume a zero entropy change for such an interconversion. The equilibrium constant was found to be insensitive to change in nitrile concentration over a small range (0.30 to 0.39 M ) , but a decrease in the base concentration from 1.0 M to 0.30 111 caused a small but significant change in K (last entry in Table I). The constant is also somewhat solvent dependent, the values for K of 1.26 (methanol, 79.9') and 1.41 (dimethyl sulfoxide, 25.2') being obtained. Since the properties of the medium change with temperature, and the equilibrium constant is slightly dependent on base concentration and solvent, these changes will be reflected in the thermodynamic parameters. In a study of the present type, it is necessary to show that the product ratio is not affected by conversion of any intermediates to product upon q11enching.l~ This can be accomplished most simply by providing evidence that the total concentration of intermediates is negligible. In an investigation of the relative rates of base-catalyzed H-D exchange and racemization of optically active open-chain nitriles in various solvents, it mas concluded that the concentration of intermediates was insignificant, as these rates are identical.16 The rates of interconversion and H-D exchange of 4-tbutylcyclohexanecarbonitriles have been determined in t-butyl alcohol-D16 with t-butoxide as ~ata1yst.l~These rates are not rapid even in the presence of a large excess of base. The kinetic behavior of the exchange indicates that the total concentration of intermediates is negligibly small. Moreover, in Table I it is shown that the relative amount of trans isomer increases with increasing base concentration in contrast to the observation that the anionic intermediate is deuterated more rapidly to give the cis nitrile.17

1-

L

I

Possible explanations for the absence of an appreciable enthalpy change in the axial-equatorial interconversion of the cyano group are that the cylindrically symmetrical n-cloud has a small radius, or that the r-cloud is strongly polarized towards nitrogen because of its greater electronegativity. A knowledge of the conformational preferences of the isocyano and ethynyl groups should help to clarify this question, and further work along these lines is contemplated. (15) See. for example, H. E. Zimrnerrnan and T. W. Cutshall. J . A m . Chem. Soc., 80, 2893 (1958). (16) D. J. Cram, €3. Rickborn, C. A . KingRbury, and P. Haherfield. ibid., 88, 3678 (1961). (17) Unpublished results.

RICKBORN AND JENSEN

4608 Experimental

The pure cis- and trans-4-t-butylclohexanec~rbonitriles used in this study were prepared by phosphorus pentoxide dehydration of the corresponding amides.'* The experimental procedures and properties of these materials have been reported elsewhere.'9 Equilibration of the Nitriles.-Stock base solutions were prepared by dissolving sublimed potassium t-butoxide (MSA Research Corp.) in t-butyl alcohol which had been dried by distillation from Linde Molecular Sieves, type 4A. The appropriate amount of either cis- or trans-4-t-butylcyclohexanecarbonitrile then was added to this stock base, so that the final solution was 1.O M in potassium t-butoxide and 0.30 M in nitrile. Aliquots (2 ml.) of this solution in sealed Pyrex tubes were placed in the constant temperature baths, and removed after equilibrium had been established.

(18) H. H. Lau and H. Hart, J . A m . Chem. Soc.. 81,4897 (1959). (IS) B. Rickborn a n d F. R. Jensen, J . Org. Chem., 27, 4608 (1962).

VOL. 27

The time required for equilibration was determined by prior kinetic experiments under the same conditions. The samples from the equilibration a t 25" were quenched by pouring into water. The higher temperature sealed tubes were cooled by immersion in Dry Ice-acetone before the contents were poured into water. The aqueous solution was extracted three times with pentane, and the pentane was washed with water and evaporated on a steam bath. Theresidual nitrile (high recovery) was used directly for analysis. Equilibrium was attained from each isomer a t each temperature, and the individual constants shown in Table I are based on eight to fifteen determinations. Analysis of Isomer Distribution .-The cis- and trans-4t-butylcyclohexanecarbonitriles were easily separable by vapor phase chromatography; a 20-ft., 1/4-in. diethylene glycol succinate polyester column used a t 150°, 20 p.s.i., gave complete peak separation. Relative areas were determined by planimetry. No correction factor was necessary for calculation of the equilibrium constants, a~ shown by measurement of a known mixture prepared from weighed samples of the pure nitriles.

a-Carbon Isomerization i n Amide Dehydrations] BRUCERICK BORN^ AND FREDERICK R. JENSEN Department of Chemistry, University of California, Berkeley 4, California Received June 68, 1966 The extent of ol-carbon isomerization in the dehydration of cis- and trans-4-t-butylcyclohexanecarboxamidehas been determined, using phosphorus pentoxide, thionyl chloride, and phosphorus oxychloride as the dehydrating agents. No detectable isomerization wm observed with the first two reagents, but some isomerization occurred with phosphorus oxychloride.

Although many dehydrating agents have been employed for the conversion of amides to nitriles, no investigations have been reported which allow a precise determination of the extent of isomerization of the a-carbon atom. Reagents which have been used include phosphorus p e n t ~ x i d e ,phos~ phorus o~ychloride,~ thionyl ~ h l o r i d e ,phosphorus ~ pentachloride,6 sulfamic acid,' aluminum chloride,8 and various organosulfonyl chloride^.^ Of these reagents, the first three mentioned are the most commonly used, and these were employed in the present study . Kenyon and Ross'O examined the phosphorus pentoxide dehydration of optically active 2-methyl3-phenylpropanamide to the corresponding nitrile, and their results indicated that no extensive racemization occurred during this process. Although (1) Supported in part b y the National Science Foundation. (2) Present address: University of California a t Santa Barbara, Gcleta. Calif. (3) R. E. Kent and S. M. McElvain, Orp. Sun., 26, 61 (1945). (4) W. B. Reid, Jr.. a n d J . H. Hunter, J. A m . Chem. S o c . , 7 0 , 3515 (1948). ( 5 ) S. M. McElvain a n d C. L. Stevens, ibid.. 69, 2663 (1947). (6) F. F. Blicke, ibid.. 49, 2848 (1927).

(7) A. V. Kirsanov and Y. M. Zolotov. Zh. Obshch. Khim., 2 0 , 284 (1930). (8) J. F. Norris and 4 . 3 . Klemka. ,I. d m . Chem. Soc., 6 2 , 1432 (1940). (9) C. R . Stephens, E. J . Bianoo, a n d F. .J. Pilgrim, ibid.. 77, 1701 (1955). (10) J . Kenyon and W. A. Ross, J . Chem. Soc., 3407 (1951).

the acid obtained on hydrolysis of the nitrile had essentially the same rotation as the material from which the amide was prepared, the amide had been recrystallized, and the possibility of optical fractionation at this step cannot be excluded. Preparations of optically active nitriles by amide dehydration have recently been reported by Cram and coworker^^^-'^; the rotational values obtained were higher with phosphorus pentoxide than with phosphorus oxychloride, but the results pointed to some racemization in most cases. In the present study, the pure cis- and trans-4-tbutylcyclohexanecarboxamides, readily prepared from the corresponding acids,14were used. The extent of geometrical interconversion in the product nitrile gave a measure of the stereochemical integrity of the dehydration reaction.

analysis of geometrical isomers [e.g., by vapor phase chromatography (v.P.c.) methods] offers a distinct advantage over polarimetric analysis, as (11) D. .J. Cram. B. Rickborn, C. .4. Kingsbury. and P. Haberfirld. J . .4m. Chem. Soc., 83,3078 (1901). (12) D. J. Cram a n d P. Haberfield. ibid., 83, 2363 (1961). (13) D. J. Cram and P. Haberfield, ibid., 83, 2364 (19FI). (14) The authors are indebted to Mr. Jnnies Rogers for samples of these acids.