3528
COMMUNICATIONS TO THE EDITOR
oxide followed by oxidation of the diol with lead tetraacetate led to the keto aldehyde VI (go%), infrared max. 5.78, 5.87 p , and this was cyclized b y dilute base in ethanol to the unsaturated aldehyde VI1 (50yo yield); infrared max. 3.67, 5.98, 6.17 p ; n.m.r. peaks a t 0.75-1.10 6 (three methyls), 2.00 6 (C=C-CHB), 4.47 6 (doublet, one acetal proton), and 9.90 6 (CHO), ultraviolet max. 266 m p ( E 9900). Hydrolysis of the acetal function in VI1 was accomplished using 1.5% sulfuric acid in aqueous tetrahydrofuran to give synthetic helminthosporal ( I ) , m . p . 55-58', [ a ] 1 8-47.8' ~ (c 1.00, chloroform) having infrared, n.m.r., and ultraviolet spectra identical with those of the natural substance. Reduction of synthetic I with lithium aluminum hydride followed b y reaction with 3,5-dinitrobenzoyl chloride afforded a bisdinitrobenzoate, m.p. 148-149.5', alone or admixed with a sample, m . p . 1 '48-149.5', derived from natural helminthosporal. Since the absolute configuration of ( -)-carvomenthone is as shown in I1 (R1 = RZ = H), the absolute stereochemistry of helminthosporal (I) follows from the synthesis. The orientation of the formyl group a t C-2 in helminthosporal as shown in I is indicated b y n.m.r. data and the resistance of this substance to isomerization in acid solution. An alternative approach to helminthosporal consists of the reaction of the ketone I11 (X = 0) with the methylene transfer agent dimethylsulfonium methylide12 to give the oxirane V l I I (%yoyield, configuration a t C-2 tentative), subsequent treatment with zinc bromide-benzene (30Yc yield) to form a mixture of C-2 epimeric aldehydes, and conversion to V (and the C-2 epimer) with ethylene glycol under acid ~ a t a l y s i s . ' ~ (12) E. J . Corey a n d M. Chaykovsky. J . A m . Chem. SOC.,84, 887 (1962). (13) W e thank Dr. Paul d e M a y o for a n authentic sample of t e t r a h y d r o helminthosporal bis-3,5-dinitrobenzoate and for helpful discussions. This research was supported b y t h e r a t i o n a l Institutes of Health.
E. J . COREY DEPARTMEXT OF CHEMISTRY HARVARD UNIVERSITY SHIGEO NOZOE CAMBRIDGE 38, MASS. RECEIVED SEPTEMBER 28, 1963
The Rate of the Tris-(2,Z'-dipyridine)-osmium(I1)Octacyanomolybdate (V) Electron-Transfer Reaction'
Sir: The application of the temperature-jump method to the study of rapid bimolecular electron-transfer reactions requires t h a t the equilibrium constant for the reaction be of the order of unity. Consequently, the standard entropy change for the electron-transfer reaction should be fairly large, so that the change in the equilibrium produced b y the temperature jump may be readily measurable. One way in which this can be achieved is by the use of oppositely charged reactants. Halpern, Legare, and Lumry2 have successfully used this method to measure the rate of electron transfer between t r is- (4,7-dimethyl-l,10-phenanthroline) -iron(11) and hexachloroiridate(1V). We report here the results of a study of the rate of the tris-(2,2'-dipyridine)osmium (11)-octacyanomolybdate (V) reaction. The temperature-jump apparatus is of similar design to that of Czerlinski and Eigen3 and Diebler4 except that we have used a single light beam instead of a dual beam. This modification improves the signal-to-noise A Bausch and Lomb grating ratio by a factor of (1) Research performed under t h e auspices of t h e U. S. Atomic Energy Commission. (2) J . Halpern, R . J. Legare, a n d R L u m r y , J . A m . Chem. Soc., 86, 680 (1963) ( 3 ) G . Czerlinski and M . Eigen, Z . Elekfvorhem , 63,652 ( 1 9 5 9 ) . ( 4 ) H . lliebler, Ph.D. Thesis, Georg-.4ugust-University, Gottingen, Germany, 1960
Vol.
ss
monochromator equipped with a 30-w. tungsten filament lamp was used as the light source. The temperature jump of about 10' was produced b y discharging a 0.02 pf. condenser, charged to 80 kv., through the solution, and the resulting change in the equilibriuin was measured b y recording the absorbance of Os(dipy)32t a t 480 nip as a function of t h e . The values of k f and k r defined b y the equation
+ M o ( C S ) * ~e Os(dipy),3+ + Mo(CN),' ki
Os(dipy),*+
k,
were found to be 2.0 X l o y F - l set.-' and 4.0 X lo9 F-' set.-', respectively, a t l o o and ionic strength 0.50 (0.45 F KhT03 0.05 F H N 0 3 ) . These rate constants are even-somewhat higher than those found for the Fe(DMP)32+-IrC162-system under comparable conditions. The diffusion-limited rate constants for the Os(dipy)32+--;\10(CN)E3-and O ~ ( d i p y ) ~ ~ + - M o ( C Nre)~~.actions calculated from the Debye equation6 are 1 X 10'O F-' set.-' and 2 X 1O1O F-' set.-', respectively, a t 10' and zero ionic strength. The values will be somewhat lower a t the ionic strength used in these studies. I t will be seen t h a t the observed rate constants lie within one order of magnitude of the diffusioncontrolled limits, as do the rate constants obtained by Halpern, Legare, and Lumry.2 Since the observed rates are close to the diffusioncontrolled limits, the energy required to reorganize the inner and outer coordination shells of the reactants and products cannot be very large. The Xarcus theory' may be used to calculate the rate constants for the electron-exchange reactions related to this oxidation-reduction reaction, and one may then compare these calculations with the observed rates. According to this theory the rate constant for an electrontransfer reaction is given b y
+
k
=
&-(w
+mlh)lRT
(1)
where 2 is the collision frequency between two uncharged molecules in solution ( l o L 11. mole-' sec.-l), w is the work required to bring the two reactants together, and m2X is the energy required to reorganize the inner and outer coordination shells of the reactants. Substitution of k = 2.0 x 109 F-' sec.-l in eq. 1 gives m2X = 3 1 kcal. mole-' a t The energy required to reorganize the coordination shells of Osdip^)^^+ and ILZO(CN)~~prior to the electron transfer thus appear to be much lower than the values calculated for many other reactant~.~,'OSince the reorganization energies are small and probably not too different in other electron-transfer reactions involving Os(dipy)Z?+, N o ( C N ) ~ ~ -O, ~ ( d i p y ) ~ and ~ + , ~ ~ o ( C N ) E(provided ~K = 1)) the rate constants for the Os(dipy)3*+--Osdip^)^^ + and &.lo (CN)s4--Mo (cN)E3- exchange reactions may be estimated from the rate constant for the Os(dipy)32+-hfo(CN)E3-reaction merely by correcting for the differences in the electrostatic work required to bring the various pairs of reactants together, and assuming the reorganization terms to be the same, for all three reactions in the first approximation. When these corrections are made, estimates of 1 X 10' F-' set.-' and 3 X l o 4 F - l set.-' are obtained for the rate constants for the O ~ ( d i p y ) ~ ~ + - O s ( d i p y ) aand "+
*
( 5 ) G. G . Hammes and J . I . Steinfeld, J . A m . C h e m S o i , 84. 4 M 9 (1962). (6) P . n e b y e , T r a n s . Elerlvorhem. Soc., 82, 265 (1942). (7) R A. Marcus, J P h y s . C h e w . , 67, 8.53 (19ci3). (8) T h e u p ier and l o n e r limits of m2X were calculated o n t h e assumption t h a t w = z m e 2 , D R and w = m a e 2 [exp ( - x R ) ] / D R , respectively The mean value of us was used i n calculating t h e electrun-exchange rates (9) N . S u t i n , A n n Reo .Yd Sci., la, 285 ( 1 9 0 2 ) . (10) B. M . Gordon, I. I.. Williams, and pi S u t i n , J .41n ('hem .Sot , 83, 2061 (1961).
Nov. 5 , 1963
COMMUNICATIONS TO
M o ( C N ) 8 4 - - M ~( CN)s3- exchange reactions a t 10' and zero ionic strength. The agreement of these estimates with the observed values of >1 X lo5 F-' sec.-' and -3 X l o 4 F-l set.-', respectively, a t essentially zero ionic strength1l-I3 is encouraging. They are in agreement with the Marcus theory and also, i t would appear, support the electrostatic calculation of the work terms, within the limit indicated, in dilute media. Acknowledgment.--\Ye wish to thank Dr. H . Diebler for valuable advice concerning the construction of the temperature-jump apparatus and Drs. R. W. Dodson and K.A. XIarcus for valuable discussions.
THE
EDITOR
These results suggest the existence of a photoequilibrium between I11 and a tautomer tentatively assumed to be the epoxyketone (IV) based on the indenone oxide analogy (I 11). Concentration of a lreshly bleached benzene solution of I11 left an oil, which rapidly crystallized on addition of methanol to give a yellow solid V 4 isomeric with 111, which was neither photo- nor thermochromic, m.p. 146-148", 238 mp ( e 16,200) and 358 (l5,200), X:Yf'l3 5.82 p (C=O). The spectral features'of the product,j its failure to form ketone derivatives, and the formation of benzoic acid on treatment with alkali support structure V. Confirmation was obtained by the synthesis shown.
X:;zz
Ph
(11) E. Eichler and A . C. Wahl, J . A m Chem Soc., 80, 414.5 (19.58). (12) M . W. Dietrich and A. C . Wahl, J . Chem. P h y s . , 38,1591 (1963) (13) R . Campion, unpublished observations
CHEMISTRY DEPARTMENT BROOKHAVES SATIONAL LABORATORY UPTON,L. I . , NEW YORK RECEIVED SEPTEMBER 11, 1963
3529
Ph
R . CAMPIOS S . PURDIE S.SUTIX
V
Photochemical Valence Tautomerizalion of 2,4,6-Triphenylpyrylium 3-Oxide
I n contrast, in the absence of hydroxylic solvents, the oily photorearrangement product of I11 retained its photochromic property and displayed absorption at h C a H u 231 mp ( e 23,000) and 2% (7100). and hC"C13 5.78 p , which was not inconsistent with its forrnu,,,itx lation as the epoxide IV. Ozonolysis of IV in ethyl acetate followed by treatment with alkaline hydrogen peroxide gave benzoic acid plus an oily acid. The latter on diazomethane esterification gave the crystalline epoxyester V I , 4 m.p. 12R-130°, Xk;,5t:'l3 5.73 p (C=O) and 12.13 (epoxide); n.m.r. 2.iO T (ArH), G.1i (OCHa), intensities 5 : 3 ; m 'e 3 12 (parent). The isolation of this epoxide strongly supports structure I V for the photobleached product. The very facile rearrangement in hydroxylic solvents of the epoxide (IV) to the pyrone (V) is unexpected. The indenone oxide (I) is stable toward alcohols but mRX
Sir : The thermochromic behavior of 2,3-diphenylindenone oxide ( I ) was recently shown to involve the reversible formation of the red benzopyrylium oxide (II).I I t was also found that. these compounds may be photochemically interconverted, the photostationary state concentration of I1 being dependent on the exciting wave lengths, but a sufficient concentration of I1 could not be ?h
?h
z
@Ph
Ph
0
-0
I
I1
Ph
Ph
I
I
Ph I
I
COOCHs
CH300C
VI
0 VI1 ?h
I11 IV obtained to permit its isolation.2 The recent reportd of the preparation of the stable 2,4,6-triphenylpyrylium oxide (111) has now permitted a more direct study of the photochemistry of these interesting dipolar molecules. A deep red acetonitrile solution of 111, prepared 1)ya modification of the reported p r o c e d ~ r e was , ~ found to be bleached to a pale yellow by irradiation with light from a 500-watt projector lgmp filtered to remove wave lengths shorter than 4500 A . Irradiation of the resulting solution with 3200-3030 8, light from a B-HG high pressure mercury arc produced an instantaneous recoloration which is attributed to the reformation of I11 by the reappearance of characteristic absorption a t 31 1 nip, 324 nip, and 6.52 p , After the intensities of the 31 1 and 524 m p peaks had increased to about 2iY0 of their original values little further change occurred on prolonged irradiation. Re-irradiation with visible light (>4500 k . )again bleached the solution and the process could be repeated. ( 1 ) E . F. Ullman and J E Milks, J A m . C h r m Soc , 84, 1315 (1YG2). ( 2 ) E . F Ullman and J E hIilks, unpublished observation ( 3 ) G Suld and C .C. Price, J A m C k e m Soc , 83, 1770 (19G1). 84, 2094 (19621
P hw : h h
ph*Phh P
0 0 VI11 IX undergoes a related rearrangement to VI1 in strong acidsfi Similarly, a compound tentatively formulated by Dilthey as the keto oxide (VIII), but which by virtue of its photochromic properties is probably the epoxide (IX), can be recrystallized unchanged from methanol but rearranges on strong heating, irradiation, or acid treatment to tetraphenyl-2-pyr0ne.~ I n contrast, the epoxyketone (IV) is converted in methanol to V in fair yield (>50y0)in several minutes at room temperature. The reaction is not inhibited by triethylamine but is strongly catalyzed by hydroxylic solvents. (4) Satisfactory combustion analyses were nbtained ( 5 ) Compare 3 , l i - d i p h e n y l - 2 - p y r ~ , n e . AF,F:"" 215 m p (e 17,800), 300 (28,800), A$!:: 5 8 4 p . ( R H. Wiley. C . H Jarboe, and I.' i K Hayes, J A m Chrrn .TOG, 79, 2002 (1937) 1. ( G ) E Weitz and A. Scheffer, Bpi., 64, 2327 (19211, C F H Allen and J W Gates, J r , J A m . Chein. Soc., 6 6 , 1230 (1943). (7) I< Putter and W Dilthey, J piaki C h r i n . , 119, 183 (1937) 160, 40 (1938) We are indebted to Professor P Yates for bringing this reference t o our attention.
