EFFECTOB SOLVEKT os Z-YALL-ES
July 20, 1901
All fractions contained sulfur dioxide which could be removed by molecular sieve 4X b u t not without introducing small amounts of other light-absorbing impurities. Fractions C, D, E were combined and refluxed with 5.6 g. of potassium hydroxide for 4 hours. ( A and B mere saved for combination with the next run.) T h e TFP was rapidly distilled aqd then fractionated. The follorving fractions were obtained. Fraction
1 cm. cell O.D. = 1 B.P.,OC.
VOI.,CC.
104 B 104-108 C 106-106. .i D 106.5 Contains some mater.
A
.50 50 617
90
7 9 4 . 5 ~
1.3204 1.3191 1.3188 1.3188
A, A. A.
Fractions C and D are eminently satisfactory as solvents for spectroscopic purposes and have been used in thin cells as low as the Cary 14 limit of 1865 A. with success. Spectra.--9bsorption curves were taken with a Cary automatic recording spectrophotometer, either model 11 or model 14. T h e maxima were in general measured by running over the maximal absorption at the slowest speed three to four times and then averaging the maxima thus obtained. The maxima could normally be duplicated to a t least 1 6 b. for the model 11 and to 4 ~ -4. 3 for the model 14. The solutions were made u p immediately before use and a 1-cm. cell was used for all the measurements, except for some of the data listed in Table \-. T h e data for the constancy of the spectroscopic properties of cyclopentadecanone are given in Table V. Air Oxidation.-Water TV%S used as the solvent for all of the cycloalkanones except cyclopentadecanone, for which isooctane was used. Air wits bubbled through the ketone solutions in the dark and the solutions were periodically examined spectroscopically, usually at intervals of 8 to 10 hours. ( T h e air used was presaturated with water.) In no case was any sign of oxidation products found, the criterion
[ COXTRIEL-TIOX FROM
THE
TABLE V CYCLOPENTADECAEONE DATA( n Solvent
Isooctane
Methanol
at a
1990 1970 1950
3147
2,2,:i,3-Tetrafluoropropanol
-+
Concn., M
0.0439 00732 ,00366 ,0413 ,00413 ,000413 .03?Il ,00351
R*-TRAXSlTlCJS)
A.
irnax, ern*.
2850 2860 2860 2830 2830 2813 =IC 30a 2786 2790
23 23 21 27 26 25 33 31
The observed optical density (10-cm. cell) was only 0.07, :md it is not possible t o determine accurately maxima under such circumstances without a different slide wire. being the amount of ultraviolet absorption between 2200 and 2600 A . In separate experiments, solutions of certain ketones were exposed to sunlight in the absence of air for about 2 months. hside from some apparent photoisomerization (a well-known reaction for such compounds), no “oxidation” was noted. hlthough these studies should not be regarded as definitive, they do suggest t h a t both nil. and l i g h t are responsible for the formation of light-absorbing materials during some of the purification procedures.
Acknowledgment.---The authors would like to thank Dr. .4. C. Haven, Jr., Jackson Laboratory, Organic Chemicals Department, E. I. du Pont de Nemours and Co., Inc., Wilmington, Del. for making the fluorinated alcohol available, and hlrs. P-K. C . Huang for supplying the procedure for the purification of the fluorinated alcohol as well as small samples of the pure solvent.
DEPARTMENT OF CHEMISTRY, THEUNIVERSITY OF VV’ISCOTSIN, MADISON6, \YISC.]
Effect of Solvent on Spectra. VI. Detection of the Solvent Effect on Molecular Conformation or Shape through 2-Values BY EDWARD h!f. KOSOWER,~” GUEY-SHUANG WU AND THEODORE s. SORENSEN RECEIVEDSEPTEMBER 6, 1960 The effect of solvent on the position of the n -+ a*-transitions of 2-fluoro-, cis-2-fluoro-4-t-butyl-, trans-2-fluoro-4-t-butyland 2-chlorocyclohexanones has been esaniiiied. Although 2-fluorocyclohexanone correlates very well with 2, the solvent polarity standard based on the charge-transfer band of 1-ethyl-4-carbomethoxypyridiniumiodide, the data for the conformationally fixed 4-t-butyl derivatives indicate that a change in the position of a conformational equilibrium is occurring with a change in solvent. The same conclusion is reached for 2-chlorocyclohexanone, for which it is also shown t h a t the equatorial isomer predominates in acetonitrile, probably because of the special mode of solvation. A qualitative theory for rationalizing the spectroscopic and thermodynamic behavior of the 8-halocyclohexanones is put forward, its chief features being the ( a ) inclusion of electrostatic repulsion (in the equatorial conformers) between the carbon-oxygen and carbon-halogen dipoles, ( b ) charge-transfer stabilization of the equatorial conformers by the interaction of the oxygen non-bonding electrons with the empty zcpper orbitals of the halogen, ( c ) stabilization of the axial conformers by a charge-transfer interaction between the a-electrons of the carbon-oxygen bond and the empty u p p e r orbitals of the halogen, ( d ) inclusion of the preference of a halogen for the equatorial position as a “steric repulsion” and ( e ) explanation of the axial shift in the ultraviolet spectra as due to a charge-transfer interaction of the ?r*-anti-bonding electron of the n + ?r*-excited state with the u p p e r orbitals on the halogens. Further data confirming the rather good correlation between 2 and electronic transitions of rigid in contrast, correlation cyclic ketones are shoirn for cyclohexrnone, isophorone and 2,2-diniethpl-3,4-dihydro[4H]4-pyrone; is not observed for 3-acetyl-8-azabic~-clo [4.4.0]dec-z-en-4-one (11, H = NCOCH3) leading t o the conclusion t h a t the compound is “folded” in non-polar solvents. The unusual spectroscopic results found for 5-methyl-2,3,4-hexatrienal in a number of so!vents suggest a n equilibrium between s-cis and s-trans forms. These cases represent applications of the solvent polarity values, Z, t o problems of molecular shape and conformation.
In the previous article,l we have shown that the relationship between the transition energies for the low-intensity jn+-x*-transition) carbonyl absorption band of certain cycloalkanones and 2( l a ) Department of Chemistry, S t a t e University of x e w York,
Long Island Center, Oyster B a y , N Y . (1) E. hI. Koson-er and G -S.K u , J . A m . Chenz. Soc., 83, 3142 (19Gl).
values (the solvent polarity measure derived from the position of the charge-transfer band of l-ethyl4-carbomethoxypyridinium iodide) is linear. Against the background of linear correlations for the ketones, c g through (210, it could be discerned that the CIE-cycloalkanone was anomalous. Its (2) E. M.Kosower, ibid., 80, 3253 (1958)
3148
EDWARD M. KOSOWER, GUEY-SIIUAIKG Wu AND T~EODORE S.SOREXSEX
Vol. E3
T A ~ LI E riBSOHPTION MAXIMA FOR
n
-C
TRANSITIONS
2-Fluorocyclohexanone's
O F 2-FLL'ORO- AND 2 - C H L O R O C ~ C L o I i E ~ . 4 N o N s s cis-2-Fluoro-4-tlrans-2-Fluoro-4-1 hutylcyclohexanone butylcyclohexarionr 2-Chlorocyclo (equatorial) (urial) hexanone Amsx €ma* A, tmnx Am&< emax
Solvent (Z)2 Xmu EmWater (94.6) 2i30 6.6' ,. .. .. '. 2796 13b TFPC(96.3)d .. .. 2775 20 2885 21 2854 24 60y0 inethanole (88.4)' .. .. .. .. .. .. 2830 13b 80% methanole (85.2)' .. .. .. .. .. .. 289 1 13' -a Methanol (83.6) 2805 5.0' 2823 s" 2915 2923 160 2830° 18Y .. .. .. .. 2R48h 24 Ethanol (i9.6) 2-Propanol (76.2)f 1803 ?ti (75.9)t .. .. .. .. '7978 19 &Butyl alcohol (71.3) .. .. .. .. , . , . 2988 30 Acetonitrile (71.3) 2883 16 2851 16 2981 17 2907 21 Dimethyl sulfoxide (70.4)d .. .. .. ,. .. .. 2893 22 2940g 18g .. .. . . Cyclohexane (60.1) .. .. .. Isooctane (60.1) 2963 18 2807 15 3020 19 3042 3G Probably low due t o 1,I-diol formation (water) or hemiacetal (metlianol). The absorption intensity of 2-fluorocyclohesanone in methanol is given for the observed optical density 5 minutes after mixing; a t 20 minutes after mixing, it had decreased t o 4.3. The observed maxima for the other Auoroketones are somewhat inaccurate because hemiacetal formation was just fast enough to distort the curves. * I t is likely that these low e-values result from addition of hydroxylic solvent t o the carbonyl group, although it is not certain t h a t solution was complete iii water. S,2,3,3-Tetrafluoropropaiiol. Derived from the correlation of n --c **-transitions of cyclohexanone with 2 as described in ref. l. e Prepared by mixing six (or eight) volumes of methanol with four (or two) volumes of water. f Ijerivctl :IS in note d. T h e linear relationship bemethanol but these lead t o tween Z and Y2gives somewhat different values (89.8 and 87.1, respectively) for 6 0 % and the same qualitative conclusions. The discrepancy indicates t h a t some care must be exercised in deriving quantitative conclusions from the correlations shown in this paper. 0 iMeasured by Kende.Ig Mentioned by C. W. Bird, R. C. Cookson and S. H. Dandegaonker, J . Chem. Soc., 3675 (1936). Measured in the usual manner? for the sample of solvent actually used in the spectroscopic determination.
behavior was rationalized as due t o a tendency to butylcyclohexanone.12 The results obtained for assume "folded" conformers in polar solvents. the conformationally mobile ketones have been We now report studies of fluoro- and chlorocyclo- supported iri an elegant manner by the preparation hexanones, axial and equatorial 2-fluoro-4-t-butyl- of the axial and equatorial isomers of 2-fluoro-,13 2cyclohexanones, several cyclohexenones, and a chloro-14 and 2-bromo-4-t-butylcycluhexanones. I5,l6 novel cumulene aldehyde, 5-rnethyl-2,3,4-hexaThe methods used to determine the ratio of trienal. The results indicate that comparison of equatorial and axial isomers have included infrared spectroscopic maxima in several solvents with absorption (examination of the carbonyl region in Z-values can provide useful and interesting infor- most cases), ultraviolet absorption (n+n*-tranmation.3 sitiori of the carbonyl group), dipole moment measurements, and optical rotatory dispersion. JX'e Results and Discussion 2-Halocyc1ohexanones.-Although it has been add in this paper the additional technique of exrecognized for some years that the configuration amination of the n-ts*-transition in a range of of acyclic a-haloketones was medi~ni-dependent~-~solvents with different poiarity as measured by only recently have Allinger and Allinger7 investi- Z-values. With care, om: can arrive a t qualitagated the variation in the conformational equi- tive conclusions concerning the change in the conlibrium of 2-bromocyclohexanone with solvent. formational equilibrium with solvent. I n addiKozima and Yarnanouchis have reported that the tion, certain anomalies in the plots of transition conformation of 2-chlorocyclohexanone varies from energies versus Z can he interpreted in terms of the solid through solution to the gaseous state. specific nevi solvstion effects. We have examined the n+n*-transition of the Additional studies of the same phenomenon have been carried out on 2-chloro- and 2-bromo-5- carbonyl band of 2-chlorocyclohexanone in eleven methylcyclohexanoneg-ll and trans-2-bromo-5-t- solvents, and t h s t of eql:atoii:il and axial 2fluoro-4-t-butyIcyclohexnnone" i i i four and five (3) T h e authors are grateful t o the Air Force Office of Scientific solvents, respectively. These are listed in Table Research for support through Contract A F 49(638)-282, t o the NaI along with a nunitrer of values for 2-fuoiocyclotional Institute of Allergy and Infectious Diseases for Grant E-1608, and t o t h e Wisconsin Alumni Research Foundation for funds granted l8 hexarioiie,.ls~ through t h e Research Committee of t h e Graduate School. (4) A preliminary account of this work was presented a t the Symposium on hfolecular Structiire and Spectroscopy, June 13, 1960, Columbus, Ohio. (5) I. h-agakama. cl a i . , J . Chem. P h y s . , 20, 1720 (1982). (6) L. J. Bellamy and R . L. Williams, J . Chem. Soc., 4293 (1957). (7) J. -4llinger and N. L. Allinger, Talrahcdron. 2, 64 (195R). (8) K. Kozima and Y. Yamanouchi, J . A m . Chem. Soc., 81, 4159
(1659). (9) C. Djerassi and L. E. Geller, Tclrahcdron, 3, 319 (1958). (10) C . Djerassi, L. E. Geller and E. J. Eisenbraun, J. Orp. Chem., 26, 1 (1960). (11) N. L. Allingcr, J. Allinger, L. E. Geller and C. Djerassi, ibid., 26, 6 (1960).
(12) C. DjPrassi, E. J Warawa, R . is. Wdff and E. J. Eisenhraun, ibid., 25, 917 (1960). (13) N. L. Allingrr and 11. hl. Blatter, J . .Im. Ckeni. Soc., 83, in press (1501). (14) N. L. Allinger, j. Allinger, L. A. Preilierg, K. F. Czaja and N. A. LeBel, ibid., 82, 5876 (1QFO). (15) N. L. Allinger and J , Allinger, ibid., 80, 5476 (1958). (16) X . L. Alliriger, J. Allinger and N. A. LeBel, ibid., 82, 2926 (1960). (17) Professor N. L. Allinger, Dept. of Chemistry, Wayne S t a t e University. generously provided us with samples of both of these ketones. Their preparation and properties are described b y Professor Allinger elsewhere. 1 1
EFFECT OF SOLVENT ON Z-VALUES
July 20, 1961 n+n'
3140
TRANSITIONS
Cydoheione
96
94
98
100
102
104
106
E &cal./rn ole 1.
Fig. ET versus Z for cis- and trans-2-fluoro-4-l-but).i cyclohexanone (the points for 2-fluor0 cyclohexanone arc iiidicated by a A ) : ET(&) = 0.115262 91.830; &(trans, = 0.120832 87.195.
+
+
in 2-halocyclohexanones from fluorine to bromine is presented a t the end of this section. A linear relationship for the n+a*-transition energies of 2-chlorocyclohexanone with Z is observed (Fig. 1) only from isoiictane to methanol, after which the slope becomes far too great for an n-+a*-transition.i~21Such a change indicates the incursion of significant amounts of another species which absorbs a t appreciably shorter wave lengths. In agreement with a previous qualitative conclusion based on the inversion of a Cotton effect curve for an optically active chloro ketone,l0 the new species is supposed to be the equatorial isomer of 2-chlorocyclohexanone. T h e linear relationship from methanol to isooctane suggests that the chief species present in these solvents is the axial isomer of 2-chlorocyclohexanone. It is likely that the most important isomer present in water is the equatorial isomer. Thus, comparison of the transition energies (carbonyl band positions) with 2 very quickly allows a qualitative conclusion as to the range of solvents in which the axial isomer will predominate, the range in which there is a mixture present in solution, and those solvents which are likely to contain mostly equatorial isomer. In this connection, one may now point to the positions of the maximum in the solvents acetonitrile and dimethyl sulfoxide. Surprisingly, the maxima for 2-chlorocyclohexanone in dimethyl sulfoxide and acetonitrile were, with respect to their positions in the ET versus Z plot, far from their expected places. Their location suggeste.d that the major component in these solvents might be the equatorial isomer. In acetonitrile, a comparison of the infrared carbonyl maximum for the 2-chloro ketone with that of cyclohexanone was in complete accord with this conclusion, whereas a similar infrared comparison in isooctane confirmed the axial character of the 2-chlorocyclohexanone present in that nonpolar solvent (Table 11). 2 2 (21) E. & Kosower, I. J . A m . Chrm. Soc., 80, 3261 (1958). (22) It is clear that the macroscopic parameter, the dielectric constant, does not reflect the polarity of a solvent in the cybolaclic
3150
EDWARD M. KOSOWER, GUEY-SHUANG Wu
AND
THEODORE S. SORENSEX
Vol. 83
TABLE I1
equatorial 2-fluorocyclohexanone. Although calis not of great value in such cases because CTCI O H E X A S O N E the effective dielectric constant is not known and the uc o , cm.-l assumption of point dipoles is really not valid, 2-Chlorocyclowe shall adopt here the estiniate of 2.0 kcal./molel9 Solvent Cyclohexanone hexanone for the repulsion. Xow, this repulsion should apIsooctane 1724-1i2ia 1727 pear speciroscopically in the form of a reduced .icetorlitrile lilZa 1T27 The carbonyl maximum in hexane is listed by Bellarrlg transition energy for the n-ta*-transition of the :1nd n‘illiamsZ4 from work of Josien and Lascombe26 as 1726 carbonyl group. However, the equatorial 2cm. while that measured2* in acetonitrile is reported as fluorocyclohexanone absorbs a t shorter wave lengths 1709 than cyclohexanone. There nmst then be present The most likely explanation for the preference a factor which prevents the electron from being for equatorial conformation in acetonitrile may excited, and that factor would also stabilize the be in the stereochemistry of solvation. In contrast equatorial conformer. iVe propose that this factor to the hydrogen bonding solvents, which are pri- is a charge-transfer interaction between the oxygeii marily acceptors through the positively charged non-bonding electrons (sp?)l and the empty 3s hydrogen, acetonitrile may act as a donor. (It and/or 3p orbitals of fluorine. If this factor is the f o r m strong complexes with nietal ions, for ex- only one which affects the position of the n--tn*ample.) Examination of models shows t h a t only transition, it must be evaluated as cu. 2.6 kcal.1’ the equatorial conformation is readily solvated a t mole. Further, if such a charge-transfer interaction the carbonyl carbon by the non-bonding pair of is present in the eyuaiorial conformer, it must also the nitrogen in acetonitrile. A similar explana- be present in the axial conformer. with its contribution diminished by t h c ?iigher ionization potention may hold for dimethyl sulfoxide. The maximuin found for 2-chlorocyclohexanone tial of the a-electrons. IYe can arbitrarily assign in 2,2,3,3-tetrafluoropropanolwas unexpected, this interaction as X3yc of the OSp*-F,.,interaction, since the fluoro-alcohol is highly polar.‘ The in- i.e., 0.9 kcal.,/mole. The parallel interaction for frared data obtained for cyclohexanone and 2- the chloro- and bromoketones i s taken as 6GG4, chlorocyclohexanone in this solvent were unfortu- (probable increased orbital overlap) of the internately too complicated in the carbonyl region to action energy estiiiiated for the equatorial combe easily interpreted, but indicated, if anything, pound. The electrostatic repulsion in the chloroa n equatorial conformation for the chloroketone in and bromoketones is also taken as 2.0 kcal.,/mole this solvent. This result illustrates the difficulty since the bond moments are quite close to that of in applying empirical solvent polarity values t o the carbon-fluorine. 2 5 ‘ To sunimarize our approach, we believe that the analysis of spectroscopic data. Nevertheless, as long as one is aware of the limitations and bears electrostatic repulsion in the equatorial conformers in mind what the Z-values actually represent in between the carbon--oxygen and carbon--halogen terms of the interaction of a solute with the sol- dipoles should lead to a reduced n+a*-trmsition vent,2 comparisons of ET with Z can provide in energy. The discrepancy between this expectation formation on the behavior of molecules in solution. and the experimental value (Table 111) for the The discovery t h a t 2-fluorocyclohexanone con- equatorial conformers is best interpreted as a tains more of the equatorial conformer in nonT A B L E111 polar solvents than either the 2-chloro- or 2-bromo2 - H . ~ L O C r C L O H E X . ~ S O S EIS ~ S O S - P O L A R LIEDIA c y c l ~ h e x a n o n e s ’emphasizes ~~~~ the need for a de1 : r, tailed consideration of the behavior, both equi2 - s kcal. jniolc i Kef. librium and spectroscopic, of these haloketones. 98.11 ... The proximity of the carbon-fluorine and carbon9 8 . GD 0.58 Table I 2-F(e) oxygen dipoles must lead to a repulsion in the 94.67 - 3.44 Table I 2-F(a) I N F R A R E MAXIMA D FOR CYCLOHEXANONE ASD ~ - C H L O R O -culation ~
w g i o n ? and is thus unsuitable for correlation of kinetic and spectroscopic d a t a . Although t h e confirmation of t h e equatorial character of 2-chlorocyclohexanone in acetonitrile was not extended t o dimethyl sulfoxide, there is no reason t o doubt t h a t t h e position of t h e n + T*transition in dimethyl sulfoxide re%ects t h e presence of equatorial conformer I n this connection, it is of interest t o refer t o t h e report of Allinger and Allinger?a on ultraviolet d a t a for t h e n + a*-transition o f 2-brnmocyclo8ctanone A plot of t h e d a t a against Z reveals t h a t the “slolie” determined by the d a t a in heptane and Y3Y0 ethanol i 4 c.u 0 0 9 , not very different from t h a t derived from t h e data for either c i s - or Iroiis-2-bromo--l-l-but lohexanonel: plotted axsinst Z: i 0 . . cu. 0.08. T h e point for 2ocyclooctanone in dimethyl sulfoxide then suggests a marked change in conformer composition t o one in which t h e major component is “equatorial,” i e’ , absorbs a t considerably shorter wave lengths in t h e ultraviolet than t h e axial isomer (or mixture of conformersza) present in heptane. T h e maximum for cyclodctanone itself in dimethyl sulfoxide reported in ref, 23 is in error, t h e value for a carefully purified sample of cyclooctanone in pure dimethyl sulfoxide being Xmax 2871 A,, cmaX 17. ’ ) J. Allinger and S . L. Allinger, J , A m . Chent. S o c . , 81, 573(i )I.
k ) I,. J. Aellamy and R. T,. \Villiams, 7‘i.niis. F i i ? ~ d ( i >.Yoi.., ’ 5 6 , 14 (1 9.59). f.5) XI. L. Josien and 1.1,;rscombe. J . C l i e m l’hys , 5 2 , 11i2 I 1
Z-Cl(e) 2-Cl(a) 5-Brie) 2-Br(a)
100.0 OR .4
1.8s --4.7
98.3
r),2
90 7
-7.4
14
11 1.i 1 ,*I
charge-transfer interaction between the oxygen non-bonding electrons aud the upper orbitals of the halogen. A lesser interaction between the T electrons and the halogen in ?he axial case must also be present. These interactions, which, added to the “steric repulsion” of axial halogen, constitute the factors that determine the relative therrnodynaniic stability of the ha!oc~clohexanones, are larger than the “induced dipole interaction” or the “van der Waals repulsion” estimated by X1linger, et uZ.,’~ and are listed in Table IV. S o simple relationship between the halogen electron affinity and the extent of interaction can be ex12.k~) R J \Xr T.e Fevre, “Di]mle Moments,” Alethiien A C o . , I,td., 1%:;. p 1:: 5.
!.IInrl:,11,
EFFECT OF SOLVENT ON 2-VALUES
July 20, 1961 TABLE I\‘ COSTRIBUTIOSS TO a4XIAL/EQGATORIAL Interaction Dipole-dipole Charge-transfer
RATIOO F
CYCLOHEXANONES‘~~ (e)F (a)F (e)CI (a)Cl 2.0 -2.0 2.6 3.9
-
(e)Br -2.0 2.2
%HALOa(Br)
Ospaxpmd
0.9
Charge-transfer
2.6
1.4
T-Xp,s,d
“Steric repulsion” NET
0.6
-0.4 0.5
-411 figures are in kcal./mole. destabilization.
1.9
-0.4 2.2
0.2
-0.4 1.0
* Xegative sign indicates
pected because the orbital overlap is not the same for each halogen. Inspection of the net interaction energies for each conformational pair shows that the proper trend in axial/equatorial ratio is given. Given the arbitrary nature of the numbers used, Table IV should be regarded as supplying a possible rationalization of the observed axial/equatorial ratios, which are themselves only known in an approximate way. A more easily demonstrated point concerns the origin of the shift to longer wave lengths found for axial 2-halocyclohexanones. Although a canmon explanation involves supFosed contributions from the resonance f ~ r r n l I~, ~presumably ?~ as a way of favoring the excited state, the accepted theoretical picture of low intensity, n+a*-transitions of carbonyl groupsz7 suggests that such electron-supply would raise rather than decrease transition energy. A similar objection holds for the related proposal that a bromine 4p(filled) orbital may overlap with the ir*-anti-bonding orbital. 2 3 The transition involves transfer of an electron from a non-bonding orbital on oxygen into a n*-anti-bonding orbital between the carbon and the oxygen. Any electron supply to this region of the molecule should hinder the excitation. It follows that a mechanism must
1
3151
ra~si.~OFurthermore, the general status of hyperconjugation (a previously used explanationz6) as a mode of electron-supply is in serious question even if no other objection could be voiced. There are many atoms (or groups) which can accept electrons into upper orbitals, and further interesting observations may be possible with molecules like substituted thiocyclohexanone. Cyc1ohexenones.-Previous studies on the -x*and n-ta*-transitions of bicyclic a,Punsaturated ketones I1 (H = CH2, NCH3, and N(CH3)2+)3 2 have shown that a good correlation is found between transition energies and 2. The correlation permitted the calculation of excited state dipole moments and the conclusion that the shift in the *a*-transition from H = CH2 to N(CH3)2+ was primarily electrostatic. In order to confirm that cyclohexenones would show solvent-sensitive light absorptions which could be correlated with 2, we have now examined cyclohexenone (111),isophorone (IV) and 5,5-dimethyl-4oxa-2-cyclohexenone (2,2-dirnethyl-3,4-dihydro[4H\-4-pyrone) (V) . 3 3
m
0
I1
6 I11
CH3
CH, IV
CH, V
The a+a*-transitions for 111, I V and V are given in Table V, while the low-intensity n+a*transitions are given in Table VI. In the case of the pyrone V, the n-+a*-transition is obscured by the vir*-transition in all solvents but isooctane. The data in Table V are plotted against 2 in Fig. 3 , and that of Table VI (treated in the same manner) summarized a t the end of the Experimental section.
exist for removing charge, and one is led quite TABLE I’ naturally to the idea that the electron in the a*P + T”-TKASSITIOSS OF CYCLOHEXES orbital should have a low ionization potential and I11 IV Amax. Am.,. should therefore be an excellent donor to upper (unSolvent (Z)2 A. emsx A. fnisa filled) orbitals on the neighboring axial halogen. TFP (96.3) 2416 14,200 The trend from fluorine to chlorine to bromine is Water (94.0) 2300 10,000 2422 13,100 2693 9800 clearly marked in the axial cases listed in Table Methanol (83.6) 2215 10,200 23.53 13.200 2635 9800 Ethanol, 95% (81.2) 2248 10,000 .. .,,, 111, and may be attributed to the increasing elec2-Propanol (76.8)” 2233 9.600 2838 12 ,000 .. .. tron affinity. Again, because of the variation in Acetonitrile (71.3) 2220 10.000 2,314 13,600 2588 9800 distances (and even in anglesz8), one should not IsoBctane (60.1) 2172 12,000 2151 14 200 2623 9600 expect a smooth correlation with size. a Measured for the solvent used. Not only is the foregoing formulation satisfactory for understanding the “axial shift” but it The solvent sensitivities found for the transimakes clear why the imperfectly “axial” bromo- tions of 111, IV, and V are in accord with those and chlorocamphors show the shift29as well as the found for other ketones, collected in Table 1711. imperfectly axial bromoketone reported by Dje(26) N. J. Leonard and F. H. Owens, J. A m . Chem. SOC.,80, 6041 (1958). (27) J. W. Sidman, Chem. Reus., 6 8 , 689 (1958). (28) V. A. Atkinson and 0. Hassel, A d a Chem. S r n n d . , 13, 1737 (1959). (29) R . C. Cookson, J . C h e m . S O L ,282 ( 1 9 2 )
(30) C. Djerassi, N. Finch and R . Mauli, J . A m . Ciiem. Soc., 81, 4997 (1959). (31) C f . t h e summary in E. M . Kosower and J. A . Skorcz, ihid , 8 2 , 2195 (1960). (32) E. M. Kosower and D . C. Remy, Telvaiiedrmz, 6, 281 (1959). (33) Convenient procedures for t h e preparation of V and VI will be reported elsewhere; T.S. Sorensen, unpublished results.
T-R' TRANSITIONS
T IN r - r c BICYCLIC TRANSITIONS KETONES
9Ot
Fig. ET (transition energies) z'eyszLs 2 (solvent polarity values) for the high intensity band of cyclohexenone, isophorone and 2,2-dirnethyl-3,-~-dihydro I411 -4-pyrone; ET(111) = -0.201462 143.56; ET(IV) = -0.243822 141.28; ET(V) = -0.202232 125.27.
+
+
TALIJ.E 1.1 ~*-Tr~4xsrrross OF CTCLONBXENOSES
v
IV
111
L*z,
Xm a* >
Solvent (Z) 'I'FP (96.3) .-. \I ater (94.6) Xethanol (513.6) IStha?lol, 95yc (81.2) ?-Propanol (iB.S)* .icetonitrile (71.3) Isooctane (60.1)
A.
a.
€mar
XI",,,
ClnHx
2920= cIla 51 2939' 09' 34 3105 47 36 .. .. 33 3154 41 30 3211 38 27 3367 31
3072 3192 3197 3229 3292 3420
A. , . .. ..
.. ..
..
tmax
..
.. .. ..
SESSITIVITIES
..
TABLE 1-11 O F KEroxB
ELECTROXIC
TRANSI-
TIONS ET = r d 4-b Compound
Cycloalkanones, C,) n = .1 1c = 5 = 6-10 Jlesitj-1 oxirlc
=
Transition
m
n+rY
0.08 . -')i j 1 Ir ) , 1.5 - .18 . '-) ( I - ,L'o . ::ti - ,211 .27 - .2-i
Ir.
I-I
=
CIT,
n
+
1:
4 T
71
,,
!I
!!I
Ref.
0
--+
1,
__L
-
?r* p
Ti* 77 ti 71
;-
T A
ir
-+
T*
n
-+
T*
1 1 1
-, . .a 32 32 h h
. .b .31 . .b T. T--7I* -.20 . .h G. S. IVu, unpublished results on pure mesityl oxide. t Present work. I \'
7r+
7r*
11
77%
--t
117
1
,
118
119
m d acetone i n which the transition energy is changed by 9.1 and 4.1 kcal./rnole. respectivel;;. l
