802
THOMAS J. KATZ,WILLIAMH. REINMUTH AND DONALD E. SMITH
Vol. 84
which proceed by way of free-radical intermediates be conducted except in slightly acid or neutral are also subject to polar influences.32 Without solution, since the trifluoromethyl olefin is extrafurther data, such as reaction kinetics and salt ordinarily sensitive to Michael addition. In the and solvent effects on the rate of reaction, a radical presence of an acetic acid-sodium acetate buffer, pathway involving transfer of a hydrogen atom or even of a methoxyacetate-methoxyacetic acid cannot be eliminated a t this time. Regardless of buffer, the Michael addition of methanol to the the details of mechanism] the model system has double bond takes precedence over reduction ; apparently led to satisfactory reaction by sub- the successful experiments were carried out in the stituting an electrostatic acceleration by a tri- presence of added methoxyacetic acid or pyridinium fluoromethyl group or a carboxyl group for the perchlorate. (In mildly alkaline solution, N-alkyl catalysis by the enzyme. pyridinium salts are also unstable.) On the other Unfortunately, the available examples of the hand, strongly acid conditions (or in some cases model system are few. Several attempts to even mildly acid conditions) are not tolerated by extend the generality of the reaction failed. A the notoriously acid-sensitive dihydropyridines. number of other dihydropyridines were tried Presumably in enzyme systems, the protein can without success; the list includes N-benzyl- catalyze the reaction sufficiently to allow reduction dihydronicotinamide, 3,5-dicarboethoxy-1,4-dihy- even under very mild conditions, where the dihydropyridine, 3-cyano-N-methpl-l,4-dihydropyridropyridines are stable. dine, 3-cyano-N-phenyl-l,4-dihydropyridine, NAcknowledgments.-The authors take pleasure phenyldihydronicotinamide, N-benzyl-3,5-dicar- in acknowledging a grant (N.I.H. 5486) from the boxamyl-1,4-dihydropyridine.The sharp require- National Institutes of Health which enabled us to ments with respect to the structure of those ketones carry out this investigation. P.E.K. wishes furwhich can be reduced has already been mentioned. ther to acknowledge his support from a PostdocThese restrictions probably arise because of the toral Fellowship awarded to him by the National limitations in experimental conditions which are Science Foundation. The authors also wish to eximposed upon the reduction. The reaction cannot press their indebtedness t o Dr. John L. Graves, who carried out some of the preliminary experiments on (32) C. Walling, "Free Radicals in Solution," John Wiley and Sons, the reduction of benzoylacrylic acid. Inc., New York, N. Y.,1957, p. 38-1.
[COKTRIUCTIUN lCRORI THE
DEPARTMENT OF
CITEMISTKY, COLUMBIA UNIVERSITY, KE\V T U R K
27,
s.T.]
Electrolytic Reduction of Cyclooctatetraene BY T r ~ o n J. ~ sKATZ,WTLLIAN H. REINXUTH AND DONALD E. SMITH RECEIVED JULY 6 , 1961 The electrolltic reduction of cycloBctatetraene in 96(;" diosaiie-water is shown by oscillopolarography t o proceed by reversible addition of two electrons to the hydrocarbon followed by an irreversible pseudo-first-order reaction of the resulting dianion with the solvent. The kinetics of the latter reaction are determined. The observed behavior supports t h e conclusions based on e.s.r. and n.m.r. studies t h a t t h e geometries of the cyclooctatetraenyl anions are similar and unlike t h a t of the hydrocarbon. The significance of the observed half-wave potential is discussed in terms of t h e compressional energy required to aromatize the cyclooctatetraene molecule and the energy level of the hydrocarbon m.0. being filled. It is concluded t h a t this m.0. is non-bonding, consistent with a regular octagonal geometry for the cyclooctatetraenyl dianion.
Introduction The high degree of stabilization due to xelectron delocalization which is characteristic of carbocyclic aromatic systems and which had been sought but found absent in the hydrocarbon cyclooctatetraene is to be found in the cyclooctatetraenyl anio1is.l Since the available observations2 concerning the electrolytic reduction of cyclooctatetraene had suggested this hypothesis to us before our own experiments were begun, we submitted the polarographic behavior of cyclooctatetraene to further scrutiny. The results of this study, which are elaborated below, bear out the hypothesis and further confirm the conclusions derived by means of proton nuclear magnetic and electron spin resonance techniques. (1) T.J. K a t z , .7. A m . Chcm. S O L ,82,3784 (1960); 82,3785 (1960); 1'.J . K a t z and H . I, Strauss, J . Chcin P h y s . , 32, 1973 (1960). (2) (a) K . M. Elofson, A x a l . Chcm., 21, 917 (1949); (b) J. H. Glover and H. W. Hodgson. A n n i y s t , 77, 473 (1952); (c) L. E. Craig, I