2911
COMMUNICATIONS TO THE EDITOR
Vol. 78
loid formulated as the di-N-oxide of lupanine, may be regarded similarly, in that it has been secured by calcium peroxide oxidation of the parent amide.’ Since thermopsine is formulatedS as a diastereoisomer (VII-a,b-cis) of anagyrine, conversion of the latter structure t o the former, as an extension of the synthetic plan, was desirable. The epimerization of dL-anagyrine was accomplished by mercuric acetate dehydrogenation-leading presumably to the unisolated intermediate imine salt-followed by catalytic reduction (67, palladium-on-strontium carbonate). The final product, after being sublimed a t 150-170” (1.0 mm.), melted a t 171172’, and no depression was observed in a mixed melting point determination with this material and dl-thermopsine (m.p. 171-173”)) obtained by mixing equal parts of d- and I-thermopsine derived from natural sources. The infrared spectra of the natural and synthetic materials in solution were indistinguishable. In the same vein, it may be mentioned that the rare Lupinus alkaloid, C Y isolupani?ae,VI11 ja,b,c,d-cis), had been previously obtained8 in a like fashion from d-lupanine ( d e
cyclobutyl group for the NSR spectra. I am indebted to Dr. James N. Shoolery of Varian Associates, Palo Alto, California.
( 2 ) L. H. hfeyer, A. Saika and H . S. Gutowsky, Tms J O U R L A L , 76, 4567 (1953); Corey, et a l . , ibid.. 71, 4941 (1955). J . IT. Shoolery, papers presented a t the 124th A.C.S. Meeting, Chicago, 1953, Abstracts. p. ISM, and a t thc I ? f i t l iM e e t i n g , S e w York, 1!)54, Ahstracts, p . m h f ; nml. Chem., 26, 14rm (1954). ( 3 ) I{. :i.Gutowsky, e1 a l . , 1111sJ O U P N A I . , 74, .I809 (l!l52) ( 4 i B. 1’. h i l r y a n d I . N Shoolery. ;bin‘, 77, 3977 (1!)5,5)
( 5 ) J. H. Simons, U. S Patent 2,519,983 (1950). ( 6 ) J . D. LaZerte, L. J. Hals, T. S. Reid and G. H. Smith, ’ r H I S J O I J R K A L , 75, 4525 (1953). (7) G . V. 17. Tirrs, €5 A . Brown and ’r. S Reid. i h i d . , 76, 5078 (1953); G. V. D. Tieis, i h i d , 77, 6703, 6701 (19%). ( 8 ) 11. 0. Hrown, e1 0 1 , :bid.. 76, 1 21 ( 1 0 5 3 ) tbid I ~ ,, 77, :MIS > CC13, CH2I > COC1, CH2Br > DEPARTMENT O F CHEMISTRY USIVERSITYOF WISCOSSIS EUGESEE.V A N TA>IELES CF2CC13. CHiCl > CzF6, n-C3F; > CF3 > CFgH Additional observations made upon perMADISON, ~VISCONSIX J O H N S. BARAS >H. fluoroalkyl chlorides, bromides, and iodides, too RECEIVED APRIL 12, 1956 numerous for presentation here, lead to the following further evaluations of apparent electron-withSTERIC AND POLAR DISPLACEMENTS OF NUCLEAR drawing power: I > Br > C1> F >> CFJ > CF2Br > CF2CI > CF3. SPIN RESONANCES’ I t appears that the bulkiness of substituents Sir: such as I and CC13 has the effect of compensating Important advances in structure determination and “fingerprint” identification of chemical com- for their lesser electronegativity in producing pounds are currently being made by nuclear spin “electron withdrawal” from nearby CF2 (and CFd) resonance (NSR) spectroscopy. The gross fea- groups. Furthermore, these bulky groups still tures of NSR spectra have been shown to cor- exert strong “electron-withdrawing?’ effects, even relate with the electron-withdrawing power of when a CF2or CH2unit is interposed. These observations are qualitatively in agreenearby groups in the inolecule.:’~.‘ I wish to report that factors, apparently steric in origin, and quite ment with the concept of “strained homomorphs,’’* unrelated to the electronegativities of substituents, provided that repulsive steric interaction with are also of considerable importance in determining neighboring atoms or groups produces a net disthe positions of SSR lines in fluorocarbon deriva- placement of electrons away from the fluorine bond. tives. The d”-values reported in Table I are in atom, and presumably along the F-C parts per million, ==I 1 p.p.m., and refer to the defi- This effect, here entitled “repulsive unshielding,” nition 6 ” = 106 ( I I c I F s - ~ ~ ~ , t s . ) l IPositive ~ ~ a p s .6”- should also be observable in proton NSR spectra.’@ values (i.e., less shielding of the fluorine nucleus by Possible “mechanisms” for non-bonded repulsive its electron cloud) thus indicate greater electron interactions have been discussed with reference to withdrawing power than is shown by the perfluoro- potential barriers to internal rotation.Y I p o s t d a t e that net electron displacement a . i o a ~Jron? (1) The Chemistry of Perfluoro Ethers. I V . Presented in part fluorine (and hydrogen) nuclei rnay be aiaduced by a t the 126th A.C.S. Meeting, A-ew York, 1954; Abstrarts, p . 2711.
COMMUNICATIONS TO THE EDITOR
June 20, 1956
2915
ionization of I, promoted by lithium bromide but without nucleophilic attack by bromide ion, is precluded by a substantial rate factor between (10) H. J . Bernstein and W. G. Schneider, J. Chem. Phys., 24, 468 lithium bromide and lithium perchlorate. On the (1956), have just published XSR data for aromatic condensed ring systems. "Anomalous" peak displacements toward lesser shielding, other hand, lithium perchlorate gives rise to reacinconsistent with molecular-orbital predictions but in accord with the tion rates as high or higher than does lithium broconcept of "repulsive unshielding," were found for the hydrogen atoms mide with exo-norbornyl, p-methoxyneophyl and a t the 4,5-positions of phenanthrene and at similar hindered positions pinacolyl benzenesulfonates, substances disposed in more complex molecules. For biphenyl the repulsive interaction of these hydrogen atoms results in deviations from coplanarity; toward ionization without nucleophilic attack by 6. G. W. Wheland, "Resonance in Organic Chemistry," Wiley, N. Y . , an external nucleophile. 1955, pp. 157-162 It is suggested that nucleophilic attack of halide COKTRIBUTIOS No. 107, CENTRAL RESEARCH DEPARTMENTion on I gives rise to an actual intermediate 11, COMPANY MINNESOTA MINIXGAND MANUFACTURING ST. PAUL6, MINNESOTA GEORGE VANDYKETIERS with the electronic and geometrical structure generally ascribed to the transition state in s N 2 subRECEIVED MARCH19, 1956 stitution. From this intermediate is obtained substitution product I11 by loss of OTs- and also MERGED BIMOLECULAR SUBSTITUTION AND olefin IV, the latter by a process which disengages ELIMINATION' a /?-proton and both the incoming and outgoing Sir : nucleophiles, Y- and OTs-. Stereoelectronic con~ that preferably an axial proton While elimination accompanying bimolecular s i d e r a t i o n ~suggest nucleophilic substitution (SN2) of alkyl halides or is lost. While a solvent molecule may remove the toluenesulfonates is not well recognized in the case proton, the leaving group OTs- appears also to be of relatively non-basic nucleophiles such as halide in a position to depart with the proton, and this is ion, we have occasionally observed such elimina- the manner in which the formation of olefin is repretion, and several pertinent examples are available sented in formulas 11-IV. On the basis of this in the literature. Mechanistically, the behavior mechanism, FE would be expected to decrease with of trans-4-t-butylcyclohexyl p-toluenesulfonate (I), ascending nucleophilicity of the anion Y-, as is which has the toluenesulfonoxy group constrained actually observed and summarized in Table I for the series of halide and thiophenolate4 ions. The to an equatorial p ~ s i t i o nis, ~instructive. identity of FE (0.25 =k 0.01) for both (n-C4H9)4 Y'o N+Br- and Li+Br- also supports the suggested mechanism rather than some form of cyclic ciselimination mechanism.2b
repulsive interactions with neighboring groups in the molecule.
TABLE I SUBSTITUTION AND ELIMINATION O F THE 4-t-BUTYLCYCLOHEXYL TOLUENESULFONATES IN ACETONE AT 75" E F -
-wu
Salt
H
With lithium bromide in acetone a t 75', I undergoes a second order reaction, partly substitution and partly elimination to 4-t-butylcyclohexene (IV). The behavior of the cis-4-t-butylcyclohexyl p-toluenesulfonate is analogous, the total reaction rate and the fraction of elimination, FE,being ca. 7 and ca. 2 times as large, respectively. Considering possible elimination mechanisms for I, the E2 mechanism is essentially precluded by stereoelectronic consideration^.^ The sequence of S N followed ~ by E2 does not account for the observed elimination, since the intermediate bromide proves to be too unreactive by a factor of a t least lo2. An E l elimination mechanism involving (1) Research supported in part by the National Science Foundation. (2) (a) A. L. Solomon and H. C. Thomas, THISJOURNAL, 72, 2028 (1950); (b) R. P. Holysz, ibid., 76, 4432 (1953); ( c ) D. J. Cram and F. A . Elhafez, ibid, 7 6 , 28 (1954). (3) S. Winstein and N. J. Holness, ibid., 77, 5562 (1955).
Rel. ratea
nb
Irans
c*s
ROTs
ROTs
1 3.04 0.30 LiCl BurNBr 10 3.89 .25 5 3.89 .25 0.57 LiBr NaI 5 5 04 ca .10 .41 High ca. 0 .48 NaSC6HsCs4 Nucleophilic constants of Total second order rate. the anions [C. G. Swain and C. B. Scott, THISJOURNAL, 75, 141 (1953)l. In 90% EtOH at 25".
While the scope of the present mechanism for elimination is not yet clear, i t should further our understanding of certain cis- or non-stereospecific eliminations and certain competitions between substitution and elimination. The proposed mechanism need not be restricted to cases of elimination of HX, and i t can be generalized to elimination of ZX, where Z is also halogen, etc. DEPARTMENT OF CHEMISTRY OF CALIFORNIA UNIVERSITY Los ANGELES24, CALIFORNIA RECEIVED MAY 7, 1956
S. WINSTEIN D. DARWISH N.J. HOLNESS
(4) E. L. Eliel and R. S. Ro, Chemistry and I n d u s f v y , 251 (1956)
