7866
J. Org. Chem. 1993,58, 7865-7869
Axial/Equatorial Proportions for 2-Substituted Cyclohexanones Ernani A. Basso Departamento de Qulmica, Uniuersidade Estadual de Maringh, Caixa Postal 331,87020-900 Maringb, PR, Brazil
Carlos Kaiser and Roberto Rittner'J Instituto de Qulmica, Uniuersidade Estadual de Campinas, Caixa Postal 6154, 13081 -970 Campinas, S&o Paulo, Brazil
Joseph B. Lambert' Department of Chemistry, Northwestern Uniuersity, Euanston, Illinois 60208-3113 Received July 16, 1993.
Axial-equatorial conformational proportions have been measured for 2-substituted cyclohexanones in chloroform by the Eliel method for F,C1, Br, I, MeO, MeS, Me2N, MeSe, and Me. For the first seven of these, at least five experimentally independent measurables were used and the resulting conformational preferences appear to be accurate to within 10% . Systematic errors degraded the results for MeSe and Me. For Me2N, the conformationalpreference also was measured for the first time a t slow exchange in the low-temperature 13Cspectrum in several solvents. In chloroform, steric and polar effects contribute to the conformationalpreferences, with steric effects dominant for large groups such as I and MeS. The conformationalanalysis of six-memberedrings has provided the foundation for modern stereochemistry. Investigations of the factors that determine conformational preferences of Substituents on saturated six-membered rings (A values) have enriched our understanding of how atoms and functional groups interact with hydrocarbon fragments.2 The introduction of unsaturation into the ring has a profound effect on conformational properties. An endocyclic double bond changes the fundamental conformational family from chair to half-chair: and the interactions of substituents at the 3 and 4 positions respond to the new steric milieu. An exocyclic double bond may be illustrated by either the exomethylene group (C=CH2) or the carbonyl group (C=O), and substituents may be introduced at the 2,3, or 4 positions. We have previously studied in depth the case of the 3-substituted exomethylene ~ y s t e m .Herein ~ we provide a comprehensive study of 2-substituted cyclohexanones in an effort to understand how polar and nonpolar substituents interact with the carbonyl group. There are significant differences between the ring shapes of cyclohexane and cyclohexanone. Whereas the C-C-C angle is 1 1 1 O in cyclohexane,6 it is about 116O for the C-(CO)-C portion in cyclohexanone.6 The Cl-C2 bond length of 1.510 A and the C 2 4 3 bond length of 1.545 A in cyclohexanone compare with 1.530 A in cyclohexane.6 These differences cause the carbonyl portion of cyclohexanone to be flattened in comparison with cyclohexane. Abstract published in Aduance ACS Abstracts, December 1,1993. (1) This work was supported by the National Science Foundation (Grant No. CHE9302747), CNPq for a scholarship to E. A. Basso,and FAPESP. (2) For a review of A values, see Hirsch, J. Top. Stereochem. 1967,1, 199-222. (3) Anet, F. A. L. In The Conformational Analysis of Cyclohexenes, Cycloheradienes,and Related Hydroaromatic Compounda; Rabideau, P. W., Ed.;VCH Publiehers: New York, NY,1989; Chapter 1. (4) Lambert, J.B.; CUeman, R.R.;Taba, K. M.; Marko, D. E.;Bosch, R. J.; Xue, L. Acc. Chem. Res. 1987,20,454-468. (5) Booth, H. h o g . N u l . Magn. Reson. Spectrosc. 1969,5,149-381. (6) Eliel,,E. L.;Allinger, N. L.; Angyal, S. J.; Morrison, G. A. Conformatronal Analysis; Interscience: New York, NY, 1965.
0022-3263/93/1958-7865$04.00/0
The barrier to rotation about a C(sp3)-C(sp2)bond as in cyclohexanoneis generally lower than that about a C(sp3)C(sp3) bond. The more flattened ring and the lower barrier cause ring reversal, the process that interconverta axial and equatorial positions, to be more rapid for cyclohexanone than for cyclohexane. For this reason it is more difficult with cyclohexanones to obtain slow-exchange NMR spectra, which provide distinct resonancesfor axial and equatorial groups. The barrier typically is 10 kcal mol-l for cyclohexanes and 5 kcal mol-l for cyclohexanones? To date, there are no examples of conformational populations measured directly at slow exchange for substituted cyclohexanones, whereas the experiment is commonplace for cyclohexanes. The conformational preference of a monosubstituted cyclohexane is determined largely by the interaction of the substituent with the syn-axial protons and the carbons to which the latter are attached. For 2-substituted cyclohexanones, such interactions still are important, but in addition the interaction of the substituent with the carbonyl group must be taken into consideration, eq 1.For
4
M 0-6 0
2
a nonpolar substituent X, there are simple nonbonded interactions with the carbonyl group, which are clearly larger in the equatorial conformation. Such interactions, first recognized for alkyl groups by Robins and Walkere and later referred to as the 2-alkyl ketone effect? proved to be relatively small. Polar interactions have proved to be more important (except with alkyl substituents). The dipole of an equatorial, electronegative substituent is nearly parallel to that of the carbonyl group and results ~~
~
(7) Bucourt, R. Top. Stereochem. 1974,8, 159-224. (8) Robins, R. A.; Walker, J. J. Chem. SOC. 1951, 1789-1790. (9) AUinger, N. L.;Blatta, H. M.; Freiberg, L. A.; Karkowski, F. M. J. Am. Chem. SOC. 1966, 88,2999-3011. Also Bee, ref 6, pp 112-113.
0 1993 American Chemical Society
7866 J. Org. Chem., Vol. 58, No.27, 1993
Basso et al.
in a repulsive interaction, whereas that of an axial, electronegative substituent is approximately orthogonal to that of the carbonyl group and results in little interaction. An equatorial arrangement thus is repulsive and should decrease the dipole of the carbonyl group. Reduction in the dipolar contribution to the resonance hybrid in fact is documented by a 20-cm-1 increase in the carbonyl stretching frequency.6 The interaction between a 2 substituent and the carbonyl group also may be described in terms of the generalized anomeric effect or hyperconjugation.lo A number of orbital interactions have been suggested, such as donation of electrons from the C-X u bond or from the lone pairs on X to the u* or ?r* orbital of the carbonyl group.11J2 These interactions may be exemplified by the valence bond structures of eq 2. The orbitals are aligned for better
bx 6" A
A-
c-c
overlap in the axial conformation, so that orbital considerations favor the axial form. There have been numerous studies aimed a t determining the conformational preferences of 2-substituted cyclohexanones. The importance of solvent is clear from these studies. Irrespective of whether dipole-dipole or orbital interactions provide the dominant factor, the more-polar equatorial form is favored by increased polarity of the solvent. Methods to determine axial/equatorial equilibrium constants include isomer equilibration, dipole momenta, infrared, ultraviolet, and proton NMR spectroscopies. When the time scale of observation is relatively long, the two forms equilibrate and render measurements of distinct conformations impossible. Allinger and Allingerl3 introduced the use of the 4-tert-butyl group to bias conformational forms. A 2-X-4-tert-butylcyclohexanone exists in two forms (eq 3),whichmay be equilibrated
Table I. Axial Percentages of 2-Substituted Cyclohexanones
X F
solvent cyclohexane
C1
CHC& 1,4-dioxane cyclohexane
Br
CHC& 1,a-dioxane cyclohexane
cc4
I OCDa SCHa CH3
cc4
cc4
CHCla 1,4-dioxane hexane CHaCN CHCh CDCla
1H (6) 56" 340 24O 200 770 72' 56O
W 87O 84a
790 700 95d 2oa
1H (J) infrared 300 340 13O 23O 700 736 65O 490 36' 26b 890 46' 87a 74= 700 68' 6Q 57'
ultraviolet
63b 37b
7Of
ge
Pan,Y.-H.; Stothers, J. B. Can. J. Chem. 1967,45,2943-2963. Allinger, N. L.; Allinger, J.; Freiberg, L. A.; Czaja, R.F.; LeBel, N. A. J. Am. Chem.SOC.1960,82,6876-6882. Allinger, J.; Allinger, N. L. Tetrahedron 1968,2,64-74. Cantacuzene,J.; Jantzen,R.;Ricard, D. Tetrahedron 1972,28,717-734. e Reference 34. f Wadislaw, B.; Viertler, H.; Olivato, P. R.;Calegao, I. C. C.; Pardini, V. L.; Rittner, R.J. Chem. SOC., Perkin Trans. 2,1980,463-466.8 Abraham, R. J.; Chadwick, D. J.; Griffiths, L.; Saneassan, F. J. Am. Chem. SOC.1980, 102,5128-5130.
N, =
R-R, RaX-R,
(5)
+ N, = 1. This approach is attributed to Elie16J4and was first applied to cyclohexanones by Garbisch.16 We have gathered in Table I some previous spectroscopic determinations of conformational preferences of 2-substituted cyclohexanones. The pronounced solvent effect clearly supports the importance of a polar component to the substituent-carbonyl interaction. The objectives of this study were to apply high-field proton techniques to the problem, to use carbon-13 resonances to determine conformational preferences, and to obtain slow-exchange spectra for the direct measurement of the preferences. N,
Results by a suitable base. Provided that the tert-butyl group does not perturb the system, such equilibration experimenta can provide conformational preferences. Alternatively, a physical property may be measured separately for the two tert-butyl isomers and compared with that for the simple 2-substituted system of eq 1. The property for the simple system should be intermediate between those for the two conformational extremes of eq 1,and the precise value is determined by the mole fractions N,as in eq 4 for (4) a general property R. If the unmeasurable R, and R, for eq 1 can be approximated by the measurable values for the distinct isomers of eq 3, the conformer populations N , and Nw may be calculated from eq 5 and the fact that (10) Kirby, A. J. The Anomeric Effect and Related Stereoelectronic Effects at Oxygen; Springer-Verb Berlin, 1983; pp 20-23. (11) Eisemtein, 0.; Anh, N. T.;Jean, Y.; Devaquet, A.; CantacuzBne, J.; Salem,L. Tetrahedron 1974, 30, 1717-1723. (12) Loudet, M.; Metras, F.; Petrieeans, J.; Pfieter-Guillouzo, G. J. Mol. Struct. 1971,29, 263-276. (13)Allinger, N.L.; Allinger, J. J. Am. Chem. SOC.1918, 80, 54765480.
All the monosubstituted cyclohexanones were known compounds, but some of our syntheses were novel and all are described in the Experimental Section. For each %-substituentbut one, we prepared the monosubstituted molecule (eq 1) and the 4-tert-butyl systems in both cis and trans stereochemistries (eq 3). In some cases the two stereoisomers were obtained separately, in others as binary mixtures. It was not necessary to separate the two isomers of eq 3 in order to obtain their spectral parameters as models for those of the pure 2-axial and 2-equatorial forms. In this study, the X groups of eq 1and 3 were H (1, 121, F (2,131, C1(3,14), Br (4, 16), I(5,16), OMe (6,171, SMe (7, 181, SeMe (8, 191, NMe2 (9,201, Me (10,21), and tertBu (11, 22) (for each case, the first compound number refers to the monosubstituted systems of eq 1 and the second number to the tert-butyl systems of eq 3; the latter are further distinguished as the 2-axial or 2-equatorial isomer, i.e., 12a, 120). Spectral assignments of the 2-protons were unambiguous by inspection except for the hydrocarbons (X = Me, tert-Bu). In the latter cases assignments were assisted by HETCOR spectra (two (14) Eliel, E. L. Chem. Ind. 1959,568. (15) Garbiech, E. W., Jr. J. Am. Chem. SOC.1964, 86, 1780-1782.
J. Org. Chem., Vol. 58, No.27,1993 7867
Conformations of 2-Substituted Cyclohexanones Table 11.
NMR Data (CDCla) for Conformational Calculations
X F
c1 Br
I Me0 MeS MeSe MezN Me tert-Bu
6 ~ , ppm b
compd 13a
6~: ppm
2
4.90
130
4.96
14a 3 140 15a 4 150 16a 5 160 17a 6 170 18a 7 188 19a 8 190 20a 9 2oe 21a 10 210 22a 11 220
4.25 4.39 4.54 4.37 4.45 4.69 4.69 4.69 4.92 3.50 3.73 3.81 3.18 3.24 3.43 3.45 3.49
3.29 3.43 3.48 2.05 2.05 2.12 2.03 2.01
2.3W 2.91 3.23 2.538 2.4W 2.408 2.238 2.148 2.138
2.18 2.32 2.43 1.16 1.03 1.02 1.01 1.01 0.99
4.68
&,c Hz
Ja,d Hz
8.80 (50.809 20.39 (49.009 22.74 (48.609 8.74 16.49 20.89 9.18 13.12 21.64 9.38 10.95 21.60 9.14 17.80 21.18 9.12 12.34 20.33 9.94 10.70 20.5Of 9.5Of 16.16 20.84 6.40" 6.80" 7.8oh
4.80 17.40 19.00 8.04 14.00 18.70 8.00 10.90 19.20 8.40 9.60 19.60 6.80 15.60 18.50 8.80 9.80 18.40 8.80 6.80 13.60 18.20
a Chemical shift for the 2-proton. b Proton chemical shift for the 2-substituent.e Line width at half-height of the 2-proton resonance. d Sum of the two measured 2-3 coupling constanta. e Geminal H-F coupling. f Determined by analogue (see text). 8 Determined from the HETCOR spectrum. The CH3-H coupling.
*
dimensional lH/13Cspectra). Carbon-13 assignments were assisted by the attached proton test and are described elsewhere. Conformational preferences were determined in chloroform by the Eliel method, eq 5, in which the observable R was (1)the chemical shift of the 2-proton ( b ~ )(2) , the chemical shift of protons on the 2-substituent (ax),(3)the line width at half-height of the 2-proton resonance, used traditionally in this context because previous observations were on low-field machines, corresponding to the sum of the vicinal couplings (52,s) plus any long-range couplings (Jav), (4) the actual measured couplings (J2,3), or (5) the carbon chemicalshiftsexpressed as substituent parameters (6~). In the last case, we found that best results were obtained for carbons 1and 2, closest to the X group. Table I1 contains the proton data used for the conformational calculations. Except as noted, these figures were obtained by inspection from the 'H spectra. The 13C data have been given elsewhere.16 In each case the 2-proton resonance is the X part of an ABX spectrum. For the hydrocarbons, the 2 resonance was overlain by other aliphatic resonances. The actual resonance position was measured from the proton-carbon-13 2D spectrum (HETCOR). The COSY spectrum was used to correlate the 2-proton and the 2-methyl protons. Normally, a 2-axial proton is more deshielded than the 2-equatorial proton" (16) Baeeo, E. A.; Kaiser,C.; Rittner, R.; Lambert, J. B. Org. Magn. Reson., in press. (17) Wellman, K.M.; Bordwell,F. G. Tetrahedron Lett. 1963,17031708.
(the opposite to cyclohexanes). The 2-axial proton resonance in the cis-tert-butyl isomers was identified as a doublet of doublets (plus long-range splittings) with a wide total width because of the axial-axial splitting. The 2-equatorial resonance in the trans-tert-butyl isomers was identified as a much narrower doublet of triplets or as a poorly resolved multiplet. One anomaly that arose from the analysis was the observation that the 2-axial and -equatorial resonances were reversed in the hydrocarbons (21 and 22), the axial proton now being more shielded. Estimations were required in two cases. The cis isomer of the methylseleno compound 19e was not able to be prepared, either directly or by equilibration. The Javvalue of 20.50 Hz in Table I1 was estimated from the values for other substituents with similar electronegativity, e.g., the value of 20.33 Hz for methylthio. For the trans isomer of the dimethylamino compound 20a, the 2-proton resonance and the methyl resonances of the dimethylamino group overlapped. The chemical shift of the 2-proton resonance was obtained from the HETCOR spectrum, but the coupling constants were not available by this procedure. Again, the value of 9.50 Hz for this coupling was estimated from cases with substituents of similar electronegativities, e.g., 9.72 Hz for methylthio. We made these estimates only for the line widths at half-height (Jav in Table 11)and made no effort to estimate the actual couplings. Application of eq 5 to the data in Table I1 readily gave conformer populations, which we express in Table I11 as axial percentages. We include in this table the results of the calculations for each datum in Table 11, as well as analogous calculations from two-carbon resonances (more precisely, from chemical shift substituent parameters's). We attempted low-temperature measurements on all systems but succeeded in obtaining slow-exchange spectra only for the dimethylamino case (9). Figure 1shows the 13C spectra as a function of temperature in CF2Cl2 containing 3% CD2C12. From such spectra we were able to determine that the percentage of axial dimethylamino is about 60% in pure CF2C12,50% in CFzClz containing 3 7% CD2Cl2, and 5 5% in CHClF2.
Discussion Past work concentrated on the 2-halocyclohexanones (Table I). The large differences between the results obtained from chemical shifts and coupling constants indicated the inherent difficulties of the method. Errors in conformational proportions could arise either statistically because the Eliel method relies on small differences between numbers or systematically because the tert-butyl system has structural differences from the monosubstituted cyclohexanone. These errors should affect the chemical shift and coupling constant measurements differently, so that comparison of results from the two data can give some indication of accuracy. Precision of measurement probably generates an error of only &3% (calculated from repeated measurements of the same quantity). In the present study we endeavored to improve accuracy by developing a whole family of observables that could independently provide data R for eqs 4 and 5. As Table I11 shows, we succeeded in obtaining at least five observables for all the substituentsexcept MeSe and Me. In the case of MeSe our failure to prepare one of the tert-butyl isomers prevented direct application of the method. The single value of 92% axial conformer was obtained by estimating J a v for the missing conformer from values for
7868 J. Org. Chem., Vol. 58, No.27, 1993
Basso et al.
Figure 1. The carbon-13spectrum of 2-(N,Ndimethylamino)cyclohexanone in CFaCl2 containing39% CDoClz as a function of temperature: (top)-70 "C, (middle)-110 OC, (bottom)-118 O C . The resonance from CD2C12 is indicated by s. The large triplet at 6 120-130 is from CFzCl2. Table 111. Axial Conformer Populations of 2-Substituted Cyclohexanone8 in CDCk
F c1 Br I Me0 MeS MeSe MeaN Me
21 49 76 99 27 75
28 93
17 (Isd) 36 68 87 28 75 92 41
11 44 74 89 25 89
20
44 63 75 20 94
50
73
88 37 85
*
17 3 45*4 71*4 88*5
28+4
85*7 (92) 38 45 38 47 44*3 0 6 50 49 (26) a See footnotes a-d in Table 11. b From the 'Bc chemical ehifta of ClandC2. CMeanofallvalueaforthesubstituent. dFromthegeminal H-F coupling.
other substituentswith similar electronegativity to MeSe. We recognize that the value obtained is only an estimate and should be regarded as such in any future, more direct studies. In Table I11 we indicate the lack of confidence in the result by the parentheses in the final, averaged column. The alkyl Substituents, methyl and tert-butyl, suffered from very small chemical shift differences between the axial and equatorial 4-tert-butyl isomers. As a result, small differences between very similar numbers provided a large
spread in results, from 0 to 50% for Me. Again, we have little confidence in the average result in the final column of Table I11 and place it in parentheses. For the tertbutyl substituent, the axial and equatorial numbers are very similar and would provide inaccurate results. Moreover, there is no assurance that the trans isomer is in the chair conformation. Large deformations or changes in the structure of the trans isomer of course would render the Eliel method entirely inapplicable. For this reason we do not even include the results in Table 111. For the remaining substituents, we believe that our results are the best available for the conformational preferences of 2-substituted cyclohexanones. The average value in each case is the result of a t least five independent measurements. The errors shown on the averages are the mean deviations of the separate values from the mean. The errors average under 10% of the mean values. We calculated the deviations of the data in each column from the means in the last column. The following are the average deviation for each method from the means: SH, 5.9; SX, 3.0; J,,, 3.7; J23, 3.0; 6~1,6.9;6~2,3.2.We cannot say whether in general the 'H and C1 chemical shifts provide the least reliable results, but in this study their results deviated much more from the means than did the other measurables. The proportion of axial conformer increases from 17% for fluorine to 88% for iodine. These numbers parallel those observed by Pan and Stothers in Table I. They reflect the increased steric, polar, and electronic repulsion of the larger halogens with the carbonyl group in the equatorial form. Straight steric (nonbonded) effects clearly are paramount, as the larger but less-polar iodine atom has the largest proportion of axial conformer (it should be borne in mind that the 2-axial position should be viewed as less sterically demanding than the 2-equatorial position). Similarly, methoxyl is 28% axial, dimethylamino is 42 %, and methylthio is 85%. Size again is of prime importance, as the larger but less-polar sulfur group exhibits the largest axial proportion. We succeeded in obtaining slow exchange spectra only from the I3C spectra of the dimethylamino system. The melting point of chloroform does not permit such experiments, so that comparisons may not be made with the conformational proportions of Table 111. The solvent of lowest polarity, pure CF2Cl2, provides the closest approach to a gas-phase conformational preference, 60 % Increased solvent polarity stabilizes the more-polar equatorial form, lowering the axial percentage to 50 % in CFzClz containing 3% CD2C12 and to 5 % in CHClF2. These observations confirm those of Pan and Stothers with halogen substituents (Table I), but with a new substituent (Me2N) and for the first time by direct measurement at slow exchange.
.
Experimental Section Proton NMR spectra were recorded on Bruker Model AW-80 and Varian Models XLA-400 and Gemini 300 spectrometers. Carbon-13NMR spectra were recorded on Varian models XL100,XLA-400,and Gemini 300 spectrometers. Low-temperature 1% spectra were recorded on the XLA-400. Infrared spectra were recorded on a Perkin-Elmer Model 300B spectrophotometer. Cyclohexanone (l),2-chlorocyclohexanone (3), 2-methoxycyclo-
Conformations of 2-Substituted Cyclohexanones hexanone (6), 2-methylcyclohexanone (lo), and 4-tert-butylcyclohexanone (12) were commercially available from Aldrich." 2-Fluorocyclohexanone (2). Cyclohexene oxide was ringopened to 2-fluorocyclohexanol by treatment with KHF2, and the alcohol was oxidized to the ketone with Cr03.19~20 2-Bromocyclohexanone (4) was prepared by the addition of Br2 to cyclohexanone.2l 2-Iodocyclohexanone (5) was prepared by the addition of 1 2 to the cyclohexanone enolate (from cyclohexanone and lithium diisopropylamide)?2 2-(Methylthio)cyclohexanone (7) was prepared by the addition of methyl methanethiosulfonate to the cyclohexanone eno1ate.m 2-(Methylseleno)cyclohexanone (8) was prepared in two steps from the cyclohexanone enolate, fiist by the addition of elementalseleniumto form the selenide and then by the addition of methyl iodide.u 2-(N~-Dimethylamino)cyclohexanone (9) was prepared by treatment of 2-bromocyclohexanone(4)with dimethylamine.26 2-tert-Butylcyclohexanone(11).*z8 Cyclohexanone was converted to 1-(trimethylsi1oxy)cyclohexene by treatment with lithium diisopropylamide and chlorotrimethylsilane. The enol ether was treated with T i c 4 and tert-butyl chloride to produce the ketone. 4-tert-Butyl-2-chlorocyclohexanone (14)29was obtained as a mixture of cis and trans isomers by treatment of 4-tertbutylcyclohexanone (12) with Clz. 4-tert-Butyl-2-bromocyclohexanone (15)1s*~ was obtained as a mixture of cis and trans isomers by treatment of 4-tertbutylcyclohexanone (12) with Brz. 4-tert-Butyl-2-fluorocyclohexanone ( 13p1was prepared in three steps from 4-tert-butyl-2-brom~clohexanone (15). Treat(18) For details of all procedures, see Basso,E. A., Ph.D. Dissertation, 1991, Universidade Estadual de Campinas, Instituto de Quhica, Campinas, SHo Paulo, Brazil, 1991. (19) Allinger, N. L.; Blatter, H. M. J.Org. Chem. 1962,27,1523-1526. (20) Wittig, G.; Mayer, Y. Chem. Ber. 1963,96, 329-341. (21) Baeyer, A. J u t u Liebigs Ann. Chem. 1894,278, 88-116. (22) Groenewegen, P.; Kallenberg, H.; Van der Gen, A. Tetrahedron Lett. 1979, 2817-2820. (23) Scholz, D. Synthesis 1983,944-945. (24) Liotta, D.; Zima, G.; Barnum, C.; Saindane, M. Tetrahedron Lett. 1980,3643-3696. (25) Mousseron, M.; Jullien, J.; Jolchine, Y. Bull. SOC.Chim. Fr. 1962, 19,757-766. (26) Heathcock, C. H.; Buse, C. T.; Kleschick, W. A.; Pirrung, M. C.; Sohn, J. E.; Lampe, J. J. Org. Chem. 1980,45,1066-1081. (27) Reatz, M. T.; Maier, W. F. Angew. Chem., Znt. Ed. Engl. 1978,17, 48-49.
(28)Reetz, M. T.; Maier, W. F.; Heimbach, H.; Athanassios, G.; Anastasious, G. Chem. Ber. 1980,113, 3734-3740. (29) Newman, M. S.;Farbman, M. D.; Hipaher, H. Organic Syntheses; Wiley: New York, 1956; Collect. Vol. 111, pp 188-190. (30) Abraham, R. J.; Griffiths, L. Tetrahedron 1981,37, 575-583.
J. Org. Chem., Vol. 58,No. 27, 1993
7869
ment of the bromo ketone with NaCN in DMSO produced trans1-cyano-4-tert-butylcyclohexene oxide. Hydrogen fluoride was generated from KHFz and BFs etherate and was allowed to react with the epoxide to produce l-cyano-2-fluoro-4-tert-butylcyclohexanol, which, on treatment with AgNOs in ",OH solution, produced the desired fluoro ketone as the cis isomer. A mixture of cis and trans isomers was obtained by epimerization with 1M HC1 in diethyl ether.a2 4-tert-Butyl-2-iodocyclohexanone (16)%was obtained as a mixture of cis and trans isomers by the addition of 12 to the enolate of 4-tert-butylcyclohexanone(from 4-tert-butylcyclohexanone (12) and lithium diisopropylamide). 4-tert-Butyl-2-methoxycyclohexanone(17)Sa was obtained in four steps from 4-tert-butylcyclohexanone(12). The carbonyl group was protected as the enamine with pyrrolidine, and the a-position was brominated. The bromine was replaced with methoxyl by treatment with NaOMe, and the ketone was liberated as a mixture of cis and trans isomers by acid hydrolysis. 4-tert-Butyl-%-(methylthio)cyclohexanone (18)= was obtained as a mixture of cis and trans isomers by the addition of methyl methanethiosulfate to the enolate of 4-tert-butylcyclohexanone (12). 4-tert-Butyld-(methylse~leno)cyclohexanone (19) was prepared as the trans isomer in two steps from the enolate of 4-tertbutylcyclohexanone (12), first by the addition of elemental selenium to form the selenide and then by quenchingwith methyl iodide.% Epimerization efforts failed, so the cis isomerwas never obtained. 4-tert-Butyl-2-(dimethylamino)cyclohexanone (20) was (15)with obtained by treatment of 4-tert-butyl-2-cyclohexanone dimethylamine.% 4-tert-Butyl-2-methylcyclohexanone (21) was obtained in three steps from 4-tert-butylcyclohexe (12).2B*9 The ketone was converted to the hydrazone with 1,l-dimethylhydrazine. Deprotonation with lithium diisopropylamide, treatment with iodomethane,and deprotedion with BF3etherate gave the desired ketone as a mixture of cis and trans isomers. 2,4-Di-tert-butylcyclohexanone(22).2G284-tert-Butylcyclohexanone (12) was converted to l-(trimethylsiloxy)-4-tertbutylcyclohexene with lithium diisopropylamide and chlorotrimethylsilane. The enol ether was treated with T i c 4 and tertbutyl chloride to produce the ketone as a mixture of cis and trans isomers. (31) Cantacuzhe, J.; Jantzen, R. Tetrahedron 1970,26, 2429-2445. (32) Moreau, P.; Casadevall, A.; Casadevall, E. Bull. Soc. Chim. Fr. 1969, 2013-2020. (33) Cantacuzbne, D.; Tordeux, M. Tetrahedron Lett. 1971, 48074810. (34) Cantacuzene, D.; Tordeux, M. Can.J. Chem. 1976,54,2759-2766. (35) Corey, E. J.; Enders, D. Tetrahedron Lett. 1976, 3-6. (36) Gawley, R. E.; Termine, E. J. Synth. Commun. 1982,12,15-18.
